Evaluation of HARQ mechanism in WiMAX

Performance Evaluation of HARQ mechanism in WiMAX

Abstract

WiMAX(Worldwide Interoperability for Microwave Access), defined by the Institute of Electrical and Electronic Engineers (IEEE) is the latest technology developed for wireless metropolitan area networks. At this juncture, many telecommunications industries are mainly concerned with the wireless transmission of data which can use various transmission modes. Data transmitted through the air is getting corrupted due to the channel interferences, background noise, etc and Hybrid Automatic Repeat Request (HARQ) mechanism is used to ensure the packets is delivered in sequence. Since HARQ is using error correction and error detection techniques, it can improve the throughput performance even when the quality of signal is poor. When the network is overloaded, some retransmissions can be delayed due to the congestion in the network. In this thesis work, a simulation model is developed with two scenarios, one with HARQ mechanism and other being improved HARQ mechanism for the applications such as video and voice over IP (VoIP). The improved version of HARQ mechanism which uses HARQ and ARQ together improves the Quality of Service (QoS) for the VoIP users are identified in my research. Therefore, the traffic load gets reduced and the throughput gets improved further.

Acknowledgements

This thesis is carried out in the wireless communication laboratory, which is situated in, Glasgow Caledonian University, United Kingdom. Accomplishing a dissertation is a major task. Achieving the task is not complete unless an individual is supported and guided.

In this moment, I would like to express heartiest gratitude to my supervisor Professor Brian G Stewart for his keen guidance, sincere help and friendly manner which inspired me to do well in the thesis.

I am grateful to lot of people for their continuous encouragement, help and trust on me. I also would like to thank my family for their moral support throughout my dissertation.

Chapter 1: Introduction

Over the past several years mobile and wired technology is the main stream for the corporate world. But the wireless technology is bringing several changes to telecommunications and data networking to cover vast area of network connectivity. Also Wireless metropolitan area networks .WiMAX, a new standard developed by IEEE focuses on the problems of outdoor wireless networks and point-to- multipoint broadband. It has many applications, including last mile connectivity for homes and businesses and provides backhaul for the wireless hot spots.

Due to congestion in wireless LANs, large numbers of packets are discarded and have to be resent again. This results in the reduction of bandwidth for the network. So IEEE 802.16 has standardized the Medium Access Control (MAC) and Physical (PHY) layer for the broadband wireless systems and introduced Hybrid automatic repeat-request (HARQ) mechanism. HARQ is an error control mechanism which combines the MAC layer automatic repeat-request (ARQ) with PHY layer to improve the reliability. In ARQ error detection information are added to the data which is to be transmitted whereas in HARQ Forward error correction bits are also added to existing Error detection bits. As a result HARQ performs better than ARQ in poor signal conditions.

When HARQ features are enabled, it will manage the schedule packet retransmissions. Each retransmission is invoked upon the failure of each packet. After transmission either acknowledgement (ACK) or negative acknowledgement (NACK) are received and analyzed. The maximum numbers of retry blocks are limited to four. If the blocks in the HARQ buffer reach their retry limit then the HARQ blocks are removed and a notification is issued. If the retransmission fails then the scheduler will reschedule the HARQ block until successful transmission or maximum numbers of allowable HARQ retransmissions are reached.

Even though the HARQ mechanism provides better performance, i have tried to implement both HARQ and ARQ parameters in my thesis, which provides improved performance than the existing HARQ mechanism.

1.1 Goal of the thesis

The primary aim of the thesis is to analyze and evaluate the performance of HARQ mechanism in the real time traffic such as VoIP and video applications. Secondary aim of the thesis is to identify a new mechanism which performs better than the proposed HARQ mechanism and to analyze the performance in the real time applications.

1.2 Objective of the thesis

  • To analyze the performance evaluation for IEEE 802.16e in real time applications
  • To review the PHY,MAC layer specifications of IEEE 802.16
  • To design a new model by deploying large number of subscribers or nodes with various traffic applications such as VoIP and video
  • To develop a new model with a combination of HARQ and ARQ mechanisms for the improvement of Quality of Service

1.3 Structure of the thesis

1.3.1 Chapter 1:

This chapter gives the brief introduction about the WiMAX with the proposed HARQ mechanism. This chapter also provides an outline about the entire thesis with the literature review along with the aims and objectives.

1.3.2 Chapter 2:

This chapter gives the detailed introduction about the evolution of basic wireless network and its types, uses and advantages. It also provides the information about the architecture of IEEE 802.16 standard.

1.3.3 Chapter 3:

The chapter 3 provides the overview of the WiMAX with its features, architecture principles, Network reference model, brief description of Physical and Medium Access control layer

1.3.4 Chapter 4:

This chapter gives the overview of HARQ mechanism with its modes of operation.

1.3.5 Chapter 5:

The chapter 5 provides the methodology of the thesis for the proposed network, the simulation model, information about the metric carried for the HARQ mechanism and the improved version of the HARQ mechanism to improve the QoS.

1.3.6 Chapter 6:

It concludes with the summary of the thesis and also provides the recommendation for the future work.

1.4 Literature Review:

There have many approaches to examine the performances of WiMAX using HARQ. According to Min-Seok Kang and Jaeshin Jang (2006), ARQ is an error detection mechanism that enables a connection to resend data at the MAC level if an error is detected. But according to Sambale, Becvar, Ulvan, (2008) implementing ARQ method will increase the delay of packets by the time spent for retransmitting the erroneous packets, decreases the network bandwidth and increases bit error rate.Moh et al (2008) states that HARQ mechanism, which is a combination of MAC and PHY layer enhancement will manage the retransmissions upon the failure of each packet.

According to H.Zheng (2007), HARQ can improve the throughput performance using the error detection correction and error techniques when the signal quality is bad. Qian Huang (2007) has developed a model to evaluate the TCP throughput over the proposed HARQ in wireless link. Srivathsan Soundararajan, Prathima Agrawal and Yihan Li (2009) analysed performance of QoS using OPNET simulator by taking different types of traffic such as VoIP and voice. For the VoIP application PCM quality is chosen and one frame was allowed to be sent in a single packet. The authors have used Best Effort mechanism. The average throughput has improved significantly because in the previous researches done by all authors proved that there is high congestion on the network and some packet losses over a considerable period of time. According to them, HARQ is used to reduce the transmission error rate and uses only part of bandwidth for the transmissions. For the Video application the authors has chosen VCR quality with frame interval time of 30 frames per second and frame size of 352x240 pixels. The above research work done by several authors lead me to analyze the performance of HARQ mechanism in WiMAX. In my initial stages, I have adopted same method as done by above mentioned authors. In my thesis, I have also considered individual characteristics like jitter, packet delay variation and retransmission rate which is not researched in detail by the authors. However, I have chosen various scenarios for analyzing the performance.

HARQ performs fast retransmissions but it fails after multiple retransmissions. In order to overcome this problem, I have also included ARQ mechanism in my thesis. If HARQ fails, then ARQ takes over from the HARQ. This will give added layer of data protection. Both HARQ and ARQ enabled offers better performance when compared to the only HARQ enabled. In my simulation work, i have used both HARQ and ARQ which yields less delay and offers better throughput.

Chapter 2: Overview of Wireless Network

2.1 Introduction to Wireless Network:

A basic wireless network consists of multiple stations communicating with radios that broadcast in either the 2.4GHz or 5GHz band and they are based on standards of IEEE 802.11. These wireless networks can be classified into two types, namely Infrastructure wireless network and Infrastructure-less network. Infrastructure mode wireless network consists of network switches which in turn connected to wireless access point through a wired connection. The infrastructure-less wireless networks does not have fixed station as the infrastructure wireless network. It is also called as ad-hoc network, because it does not rely on routers or access points.

2.2 Why wireless networks:

A wireless network offer great productivity, cross vendor interoperability, free communication over reasonable distances. The major advantages of wireless networks are :

  • Installing a wireless network can be very easy and fast which eliminates the use of cables in the walls.
  • Initial investment for the wireless network is very low when compared to the wired network. It also reduces the installation expenses and the less need for the technical support for the installation
  • Wireless network can be easily configured starting from peer- peer networks which are suitable for the limited number of users to the big infrastructure networks.
  • There is no need for the Ethernet cable for connection with computers.

2.3 Limitations of wireless network:

Even though wireless networks have more advantages over wired network it has several limitations like roaming, performance and the security. It takes less time to set-up the wireless network but it takes more time to deploy the encryption methods. Even the wireless network can be attacked by the hackers easily to steal the information. In the wireless communication the allocation of bandwidth is limited, so smaller data bits have to be transmitted to make the communication much faster. So many applications have to be reconfigured again to use the wireless networks. The wireless networking signals can be affected by interference and propagation effects [1]. For this reason the important network servers are rarely connected with wireless networks. The speed on the most wireless networks is really slow when compared to the wired networks. The newer standards like 802.11n are addressing the limitations of the throughput of wireless networks.

2.4 Types of wireless networks:

There are four types of wireless networks such as wireless LAN, WAN ,PAN and MAN.

2.4.1 Wireless Local area networks (WLAN)

WLAN network can be deployed with one or more access points connected with the wired network with the help of cables. These technologies give the workgroup members to access the applications from a single room to an entire campus. It also contains some access points which define a finite region of coverage. Example of WLAN is 802.11 standards [2] [3].

2.4.2 Wireless Personal Area Networking (WPAN)

WPANs usually used to interconnect the compatible devices near a central location. This technology mainly used in the scenarios where the connectivity between the devices facilitates data sharing for small number of individuals. For instance the data sharing between PDA and the personal computer uses WPAN. It has a typical range about 30 feet. An example of WPAN is Bluetooth [2] [3].

2.4.3 Wireless Wide area networks (WWAN)

WWANs are created through the mobile phone signals and it is maintained by specific cellular service providers. This wireless technology differs from WLAN as it enables the users to connect even outside the boundary. The connectivity is done via PC card cellular modems. It is mainly useful for the business people as they connected for communications while travelling. An example of this technology is Cellular communications [2] [3].

2.4.4 Wireless Metropolitan Area Networks (WMAN)

A wireless Metropolitan Area Network (WMAN) which is also called Wireless Local Loop (WLL) provides wider connectivity rather than the limited connectivity between the multiple locations. WMANs are mainly based in IEEE 802.16 standard can use the radio wave or infrared light to transmit the data. The best known WMAN is WiMAX, which can reach the speed over 70 Mbps over several kilometres [2] [5]

2.5 Advantages of WMAN:

The basic advantage of using WMAN technology is the cost effectiveness. This technology also supports the converged VoIP and video streaming applications. This technology is easy to establish in any infrastructure. It has also got less complexity for the quick installation for the leased lines. The WMAN technology also offers high quality of service. The most important aspect of this technology is the flexibility in the bandwidth and also it got highly secured access. So it better provides better security when compared with the other wireless networks.

2.6 Overview of 802.16 Wireless Networks:

The wired technology is replaced with WMAN with the enhancements defined in IEEE 802.16 standards. IEEE 802.16 wireless networks defined a broadband access technology known as WiMAX which contained the features like multicast polling and QoS. The first 802.16 standard was used in April 2002 which is defined as the wireless MAN air interface. It works in the range of 10-66 GHz range for the point to multipoint type of networks. IEEE 802.16 standards are mainly concerned with the air interface between a subscriber's transceiver station and a base transceiver station. [6]

2.7 Architecture of IEEE 802.16 standards

IEEE 802.16 standards are organized into three-layer architecture.

1. Physical Layer:

Physical layer which is the lowest layer specifies the modulation scheme, type of frequency band, error correction techniques, data rates and the structure of time-division multiplexing (TDM). In order to get the quicker transmission between the subscribers and the base station, IEEE 802.16 standard uses Demand Assignment Multiple Access with Time Division Multiple Access (DAMA-TDMA). DAMA is a capacity assignment technique which responds to the changes between the multiple stations. The transmission from the base station to subscribers is done in two modes namely mode A and mode B. Mode A supports the continuous transmission stream such as audio and video, whereas mode B does in burst transmission stream such as IP- based traffic [6].

2. MAC Layer:

Media access control (MAC) layer which is situated above the physical layer contains the functions including transmitting data in the frames, controlling access to shared wireless medium. The MAC protocol defines when and how the subscriber station or base station initiates the transmission. During the downstream, MAC protocol is very simple because it has got only one transmitter. In the upstream multiple subscriber stations request access which results in complex MAC protocol.

3. Convergence Layer:

Convergence layer is situated above the MAC layer provides the functions specific to the service being provided which includes digital audio multicast, digital telephony, wireless trunks in telephone networks, Asynchronous Transfer Mode (ATM) and frame relay [6].

Chapter 3: Overview of WiMAX

3.1 Introduction

WMAN is a promising broadband wireless access (BWA) technology which operates at high speed, provides higher bandwidth and gives high capacity for the multimedia services for residential applications. As this technology covers large distances many users can use this network and can eliminate the suburban and rural areas where they cannot find this wireless access. Basically WiMAX is of two types fixed and mobile. The fixed WiMAX is based on 802.16 standards and does not have one base station to transfer to other. The second one is Mobile WiMAX, based on 802.16e supports the transmission from one base station to other. For security maintenance WiMAX uses the media access control sub-layer of Transmission Control Protocol (TCP) which handles the encryption of data packets.

3.2 Features of WiMAX

WiMAX offers set of features which has lot of flexibility in terms of deployment and potential service offerings which makes it popular day by day. Some of the salient features are as follows:

3.2.1 Mobility

WiMAX offers mobility with IEEE 802.16e-2005 uses the Scalable Orthogonal Frequency Division Multiple Access (SOFDMA) in its physical layer. Especially in mobile WiMAX it supports the seamless handovers less than 50 ms for real time applications such as VoIP. The system also consists of built-in support for the power saving mechanisms which extends the battery life for the subscriber devices. The presence of flexibility key management schemes assures the security is maintained during the handover [10]

3.2.2 OFDM-based physical layer:

The physical layer (PHY) is based on orthogonal frequency division multiplexing that offers good resistance to the multipath and also allows WiMAX to operate in the Non-Line of Sight (NLOS). Orthogonal Frequency Division Multiplexing (OFDM) is now recognized as the choice for mitigating multipath for wireless broadband [10]

3.2.3 Very high peak data rates:

WiMAX can support high peak data rates such as 74 Mbps when it operates in the 20 MHz spectrum. If it uses the 10 MHz spectrum with Time Division Duplexing (TDD), the data rate is about the 25 Mbps for the downlink and 6.7 Mbps for the uplink with the downlink- uplink ratio is 3:1. The above said PHY data rates are achieved while using the 64 Quadrature Analog Modulation (QAM) [10]

3.2.4 IP- based architecture:

The WiMAX is based on reference network architecture which is based on all-IP platform. All end to end services like end- to end transport, security, mobility and QoS depends on IP-based protocols.

3.2.5 Adaptive Modulation and coding (AMC):

WiMAX supports various modulation and Forward Error Correction (FEC) coding schemes which helps the user to change the scheme based on user and frame basis. The Adaptive Modulation and Coding (AMC) is the effective mechanism to give maximum throughput even in a time- varying channel. AMC is supported at the receiver for the signal to noise ratio and interference ratio, so that each user can be provided with the highest possible data rate.

3.2.6 Link-layer retransmissions:

To enhance the reliability, WiMAX supports the automatic retransmission requests (ARQ) at the link layer. ARQ requires each transmitted packet has to be acknowledged at the receiver end. The unacknowledged packets are retransmitted under the assumption that it is lost. Also WiMAX supports the HARQ which is the combination of FEC and ARQ.

3.2.7 Support for advanced antenna techniques:

The physical layer design allows multiple antenna techniques such as spatial multiplexing, space-time coding and beam forming. All these schemes are aimed at the improvement of overall spectral efficiency and the system capacity by using multiple antennas at the transmitter or at the receiver.

3.2.8 Quality-of-Service-Support:

The WiMAX MAC layer is designed as connection oriented architecture to support the variety of applications such as voice and other multimedia services. The MAC layer supports the real time and non real-time traffic flows, variable bit rate, constant-bit rate and best-effort data traffic to support a large number of users with multiple connections depending upon the QoS requirement [9]

3.2.9 Robust Security:

By using Advanced Encryption Standard (AES) with the key-management protocol, WiMAX supports strong encryption. Also with the help of Extensible Authentication Protocol (EAP), the system offers flexible authentication architecture to help various user credentials including the username/password, smart cards and the digital certificates. The mobile WiMAX also uses the Cipher-based Message Authentication Code (CMAC) and Hashed Message Authentication Code (HMAC) based control message protection schemes [11].

3.3 WiMAX standards & Specifications

3.3.1 Fixed WiMAX (IEEE 802.16-2004):

The IEEE 802.16-2004 is the current version for the fixed WiMAX and it can be referred as IEEE 802.16 d standard. It provides a fixed line connection with an antenna mounted on a rooftop [7]. It operates between 2 and 11 GHz and the radio interface is based on OFDM with 256 carriers. OFDM gives good resistance to interference and multiple path fading [8]. The challenge for fixed WiMAX air interface is the need to set up high performance radio links capable of data rates. (Tom Carpenter, 2006)

3.3.2 Mobile WiMAX (IEEE 802.16e-2005):

This is the extended version of 802.16a standard and it uses the mobile version to help client machines get connected to the internet. The mobile WiMAX supports handover which allows the users moving vehicular from area to area to seamlessly switch over between base stations to base stations. The mobile WiMAX operates in the range of 2-6 GHz [8]

The mobile WiMAX supports both time division multiplexing [TDD] and frequency division multiplexing [FDD]. But mobile WiMAX supports TDD mode for the following reasons:

  • It enables dynamic allocation of downlink(DL) and upper link(UL) radio resources to support asymmetric traffic in Internet applications
  • A single frequency channel in DL and UL can give more flexibility for the spectrum allocation [8]

3.4 WiMAX Network Architecture Principles

The mobile WiMAX End-to-End (ETE) Network Architecture is developed on all-IP platform with the packet technology. The network design principles underlying in the network architecture are stated below:

3.4.1 4G System Characteristics

4G systems have given some new design challenges and three major ETE requirements.

1. Limited Spectrum:

Since spectrum is a constrained source, it requires number of some innovative techniques to increase the system coverage and capacity. A combination of Pico/micro/macro-cellular access nodes will imply heterogeneity in access node capabilities which must be accommodated and evolved independent of specific operator types and core networks these access nodes integrate with. [9].

2. Support for different topologies:

The factors like system cost and scale to demand has to be considered for the access networks to employ the heterogeneity of configurations and the interconnectivity technologies. So any system designed must accommodate known and future innovations without requiring significant redesign [9].

3. Service heterogeneity:

4G radio access technologies can deliver different types of IP services from narrow-band to broadband, real time and non- real time, unicast and multicast. When it is combined to accommodate various levels of user mobility, these access systems need advanced capabilities for QoS and radio resource management [9].

3.4.2 Design Principles for the WiMAX network:

The WiMAX end-to-end delay network architecture has to fulfil the following design principles:

1. IP services optimized Radio Access Service Network (ASN)

The wireless industry marching towards the OFDM multi-access and also combined with advanced & adaptive coding schemes, FEC schemes, UL/DL channel allocation, advanced antenna systems such as multiple-input and multiple-output(MIMO) and beam forming for the broadband air interface foundation to simultaneously deliver IP services like real-time voice, store and forward unicast and multicast/broadcast of multimedia [9].

2. IP-interconnected ASNs

The WiMAX network should enable or increase the flexibility while designing the network. Important aspects like flattening ASN architectures with functional autonomy should be pushed to the radio access edge has to enabled. Also the IP-based interconnectivity should enable the inherent redundancy while using the low-cost and high bandwidth wireline or any wireless backhaul interconnectivity [9]

3. Logical separation of ASN and Connectivity and Application Service Networks:

The support for access network sharing between one or more connectivity service operators has to be enabled. Support for a connectivity service operator to offer broadband IP services over access service networks deployed by two or more operators [9]

4. Network of ASNs

The future mobile broadband systems should help the heterogeneity of access technologies. The seamless access and mobility has to be open on interoperable Industry- standard IP protocols. When the access and connectivity service network lines become blur, open IP-based interfaces between radio access components and core IP service functions have to permit independent evolution and migration to future mobile broadband access technologies[9]

3.4.3 Adopting a Functional Architecture Model:

Both 2.5 G and 3.5G are using the physical system design principles based on the access and the core design principles. This approach also enables the precise characterization of system performance and also stifles innovation in access systems. The network work group has defined a functional architecture consisting of functional entities. The prominent feature of Networking Working Group (NWG) is the extensive use of IP because it helps in enabling the access for the mobile devices.

It also defines the network functionality requirements for client devices which have the standard IP protocols like Dynamic Host Configuration Protocol (DHCP), EAP and mobile IP protocols. IP connectivity is assumed between all interacting entities in the network. The mobile IP is used as the mechanism for redirection of the data as the mobile devices starting moving from one Access Service Network (ASN) to other ASN. Proxy IP provides the mobility support for the mobiles which are not capable of using Mobile IP [9]

3.5 Network Architecture

The WiMAX network architecture is designed to meet the requirements by maximizing the use of open standards and simple IP architecture. The Network Reference Model (NRM) is a logical representation of the network architecture. It serves as a framework for evaluating the performance of the radio interface. Figure 3.1 illustrates the network reference model for mobile WiMAX

The WiMAX NRM consists of three logical parts:

  • Mobile Stations (MS) - consists of all the user subscriber devices such as wireless laptops, cell phones, PDAs and also the software which is required for communication with a wireless network.
  • Network Access Provider (NAP) provides full radio functionality. NAP also includes some functions like Access Service Network (ASN), Access Service Network Gateway (ASN-GW), base stations, foreign agent (FA), QoS and policy management. ASN helps for the network discovery and selection of the subscribers preferred Connectivity Service Network (CSN). The ASN also defines logical boundary for functional interoperability with WiMAX clients. The ASN-GW acts as a server for network session and mobility management. It also provides admission control , caching of subscriber profiles and encryption keys and service flow authorization(SFA).
  • Network Service Provider (NSP) - provides IP connectivity services. It also contains the functions like connectivity service network (CSN), home agent (HA), home servers (VAAA or HAAA), IP address management, authentication, authorization, accounting and mobility and roaming between ASNs. The Connectivity service Network (CSN) is a set of network functions which provides IP connectivity services to the subscriber stations. CSN allocates IP address to the MS for the user sessions and subscriber billing. Inter- CSN tunnelling also supports roaming between NSPs. Furthermore CSN also provides gateways, interworking with networks like public switched telephone network (PSTN) and 3GPP.

The WiMAX network reference model also contains several reference points from R1 R7. These reference points are the conceptual links which connects two functional entities and also represents the bundle of protocols between the peer entities.

3.6 WiMAX Physical layer Description:

The WIMAX PHY layer defines the conversion between the data bits from the upper layer and its corresponding electrical signals transmitted over the air. The standard specifies two entities namely; Base stations (BSs) and Subscriber Stations (SSs). There is also five air interfaces for the PHY layer to provide flexibility for the service providers:

  • WirelessMAN-SC: This air interface is single-carrier modulated for the frequency bands ranging from 10-66 GHz.
  • WirelessMAN-SCa: A single-carrier modulated air interface used for the licensed bands below 11 GHz.
  • WirelessMAN-OFDM: This air interface uses orthogonal-frequency-division- multiplexing scheme and also uses 256 orthogonal carriers for the licensed bands below 11 GHz.
  • WirelessMAN-OFDMA: An OFDM scheme consists of 2048 carriers for the licensed bands below 11 GHz. Multiple Access for SSs is achieved by using 2048 carriers.
  • WirelessHUMAN: Wireless High-speed Unlicensed Metropolitan Networks. It is used for license-exempt bands below 11 GHz[9]

The first four air interfaces support both TDD and FDD, whereas the WirelessHUMAN air interface supports only the TDD. Current WiMAX systems use both the fixed and mobile versions and it is built on 802.16 specifications. The fixed version of WiMAX is based on the OFDM PHY and the mobile version of WiMAX is based on the Orthogonal Frequency-Division Multiple Access (OFDMA) PHY.

3.6.1 Description of OFDMA

An OFDM system is implemented by multiplexing a single high data rate input stream into a parallel combination of low data rate streams. The OFDM modulation technique can easily reduce the multipath interference as it uses cyclic prefixing. These parallel streams are modulated onto separate subcarriers in the frequency domain through the use of an inverse fast Fourier transform (IFFT), which enables number of sub-carriers (up to 2048) with low complexity [12]. Generally in OFDM system, resources will be available in the time domain by means of OFDM symbols, whereas for the frequency domain, it is done by sub-carriers

OFDMA is a multiple access scheme that provides multiplexing operation of data streams from the multiple users onto the sub-channels in the downlink and uplink multiple accesses with the help of uplink sub-channels. OFDMA also subdivides the available bandwidth into multiple subcarriers in the frequency domain and symbol periods in the time domain. In the each symbol period, input data stream of a specific user is divided into number of parallel streams. Each of the parallel streams is associated with a lower data rate when compared with the original user input data stream [9]

3.6.2 OFDMA Symbol Structure

An OFDMA symbol consists of three types of sub-carriers namely data, pilot and null (guard and DC). The OFDMA symbol structure is shown in Figure 3.6.1

Data subcarriers are used for the data transmissions. Pilot subcarriers are used for the synchronization and the estimation purposes. The null subcarriers are used as guard bands for the spectrum mask requirements. In typical OFDM systems, the direct current (DC) subcarrier is not modulated. Data and pilot subcarriers are grouped into various sets of subcarriers and called as sub-channels. OFDMA PHY supports sub channelization in both Uplink (UL) and Downlink (DL). The minimum frequency-time resource unit of sub-channelization is one slot, which is equal to 48 data subcarriers [9]

In mobile WiMAX there are two types of subcarrier permutations namely diversity and contiguous permutations. The diversity permutations family includes FUSC (Full Usage of the Sub-channels), PUSC (Partial Usage of the Sub-channels), OPUSC (Optional PUSC), OFUSC (Optional FUSC) and TUSC (Tile Usage of Sub-channels). The sub-channel organization of OFDMA frame is shown in Figure 3.6.2

The main advantages of these diversity permutations are averaging the inter-cell interference and minimizes the probability of using the same subcarriers in the adjacent cells. But the channel estimation is not very easy as the subcarriers are distributed over the available bandwidth [13]. The contiguous permutations family consists of Adaptive Modulation and Coding (AMC). Channel estimation is easy as the subcarriers are adjacent.

The permutation modes defined for the mobile WiMAX are:

  • For Downlink: PUSC,AMC,FUSC
  • For Uplink : PUSC and AMC

3.6.2.1 PUSC (Partial Usage of the Sub-channels)

PUSC are divided into two zones namely DL-PUSC and UL-PUSC. DL-PUSC is the default allocation method. All the DL subframes starts in the DL-PUSC zone. All the subcarriers are divided into clusters of 14 contiguous sub carriers per symbol. In this zone, pairs of pilots swap their positions on alternate symbols, averaging one in seven of the subcarriers. The sub-channels can be divided into larger groups called segments. These segments can be divided into three segments. DL-PUSC is designed to minimize the probability of using the subcarrier in adjacent cells. For the UL PUSC, the slots are defined as the sub-channel that occurs over the three symbols. The four contiguous sub-carriers are formed over three symbols and this arrangement is called tile. The first PUSC zone always uses single in single out (SISO) systems. The symbols are divided into clusters for the downlink and tiles for the uplink. The PUSC channel consists of subframes like preamble, Frame control Header (FCH), downlink map,(DL-MAP),uplink map(UL-MAP), DL bursts and UL bursts. The OFDMA TDD frame structure showing only PUSC zone is shown in Figure 3.6.3

The DL subframe starts with a preamble in which the OFDM symbol can use the three subcarrier sets. The preamble subcarrier set works on binary phase-shift keying (BPSK) modulation and contains a group of selectable sequences. All these sequences are of Pseudo-noise (PN) sequence. The preamble is very useful to synchronization and to the initial channel estimation. Next to the preamble is the Frame control Header (FCH), downlink map,(DL-MAP),uplink map(UL-MAP) which carries the broadcast messages in the PUSC mode. These broadcast messages inform the MSs about how the DL and UL bursts should be organized in time and in the frequency. In the DL every two successive symbols is grouped to form as a unit known as subchannels whereas in the UL every three successive symbols is grouped to from subchannels.

3.6.2.2 FUSC (Full Usage of the Sub-channels)

FUSC is used only in the downlinks. FUSC uses all the subcarriers and provides high degree of frequency diversity. All the subcarriers are divided into 48 groups. A sub-channel is formed by using one subcarrier from each group. FUSC also minimizes the probability of using same subcarrier in its adjacent cells. FUSC pilots can be used in both fixed and variable positions.

3.6.2.3 Adaptive Modulation and coding (AMC)

The sub channelization Scheme based on the contiguous subcarriers is called band Adaptive Modulation and coding (AMC). This zone has wider bandwidth when compared with the PUSC and FUSC. It has got contiguous sub-channels which is best suited for low mobility applications [14]. Even when the frequency diversity is lost, AMC allows the designers to exploit the multiuser diversity. The multichannel diversity gets increased when the system provides each user with a sub-channel. When allocating a band in AMC, it defines a set of nine contiguous subcarriers within an OFDMA symbol called bin. A group of four rows of bins is called physical band.

3.6.2.4 Optional PUSC (OPUSC)

OPUSC has the tile structure which has three subcarriers by three symbols. OPUSC has got nine subcarriers, the centre subcarrier is used as a pilot and remaining subcarriers is used as the data subcarriers. The UL OPUSC is same as UL PUSC except it uses the tile size which has three subcarriers wide by three symbols long.

3.6.2.5 Tile usage of sub-channels (TUSC1 and TUSC2)

These systems are available only in the DL by using adaptive antenna system (AAS). These both are optional and it is similar to DL PUSC and OPUSC and uses different equations for the assignment of subcarriers within the sub-channels [15]

3.6.3 Frame Structure of OFDMA:

The mobile WiMAX supports both the TDD and FDD, but TDD is the preferred duplexing mode for the following reasons:

  • TDD requires only a single channel for the both uplink and downlink, thus provides greater flexibility for the global spectrum allocations. FDD requires a pair of channels for the uplink and downlink.
  • TDD makes sure it has got channel reciprocity which leads to the better support for the smart antennas and for link adaptation.
  • TDD helps for the adjustment for the downlink/ uplink ratio to support asymmetric downlink/ uplink traffic whereas in the FDD both uplink and downlink are fixed with equal bandwidths.

The OFDMA frame consists of DL sub-frame and UL sub-frame. The FDD frame structure is a flexible one and comprises of movable boundary between DL and UL subframes. There is short transition gap between the DL and UL sub-frames called transmit-receive transition gap (TTG). Upon the completion of the UL sub-frame there is short gap added between this and the next DL sub-frame called receiver-transmit gap (RTG). The transition gaps are defined in terms of physical slot (PS) units. Figure 3.6.4 shows the DL and UL sub-frames separated by TTG and ending with RTG. The figure also shows the preamble, frame control header (FCH), downlink media access protocol (DL-MAP), and uplink media access protocol (UL-MAP).

In a frame, the following control information is used for optimal system operation:

  • Preamble: The preamble is used for the timing and frequency synchronization. It is also used for the channel estimation purposes as it is the first symbol of the frame. The DL sub-frame begins with one symbol which is used for the BS identification. This symbol is generated by using PN sequences and the data in the every preamble is mapped into every third subcarrier. There are no preambles in the UL except when the systems using AAS[9]
  • Frame control header (FCH): The FCH follows the preamble with a fixed location and duration. It also provides the frame configuration information such as mobile application part (MAP), message length and coding schemes. FCH contains the downlink frame prefix (DLFP). DLFP specifies the length and coding of DL-MAP and also updates the ranging allocations which occur in the subsequent UL sub-frames.
  • DL-MAP and UL-MAP: The DL-MAP and UL-MAP provides the sub-channel allocations for the DL and UL sub-frames. MAP contains the number of zones, frame number, location and the contents of all the bursts. Each of the bursts are allocated by the symbol offset, number of symbols, sub-channel offset , number of sub-channels and the power levels.
  • UL ranging: The UL ranging sub-channel is allocated to MSs for closed loop-time, frequency, bandwidth requests and power adjustments[15]
  • UL channel quality information channel (CQICH): It is assigned to get the feedback from MS about the channel-state information.
  • UL acknowledge (ACK): It is allocated to receive the DL HARQ acknowledgements.

3.6.3 Advanced PHY Layer Features:

AMC, HARQ, fast channel feedback are the most important features of the mobile WiMAX to enhance the coverage and capacity for WiMAX in mobile applications. HARQ is enabled using the N channel stop and wait protocol which gives the fast response and improves the edge coverage. A dedicated ACK channel is provided in the UL for HARQ ACK/NACK signalling. When HARQ is combined with the channel Quality Indicator (CQI) channel, adaptive modulation coding it becomes the powerful mechanism [15]. A detailed description of HARQ mechanism is given in Chapter 4.

WiMAX supports variety of modulation and coding schemes. Using CQI, mobile can give the feedback to the base station about the downlink channel quality. Depending upon the received channel quality, base station can calculate channel quality for the uplink. The base station assigns a modulation and coding scheme to increase the throughput of the user. AMC increases the overall capacity, as it allows the trade-off between the throughputs on each link. For the downlink Quadrature Phase Shift Keying (QPSK), 16 QAM and 64 QAM are mandatory and 64 QAM is optional for the uplink. In total there are 52 combinations of modulation and coding schemes are available. Fast scheduling is also important feature of WiMAX PHY layer. Fast scheduling enables the rapid response to variations of traffic and channel conditions. Mobile WiMAX uses the fast scheduling in both uplink and downlink.

3.7 WiMAX MAC layer Description:

The mobile WiMAX MAC layer provides a standard medium-independent interface between the PHY layer and the upper transport layer. The MAC layer schedules the usage of the air-link resources. MAC layer has the special packet called MAC Service Data Units (MSDUs) which makes the transmission possible over the air. In fixed and mobile WiMAX there is a presence of convergence sub-player which makes the interface possible to the upper layer protocols. WiMAX MAC has got unique features to address SS and BS. Each SS carriers 48-bit IEEE MAC address and BS carriers 48-bit Base Station ID.

3.7.1 MAC Sub-layers:

WiMAX MAC layer is divided into three sub-layers namely Service Specific Convergence Sub-layer (SSCS), Common Part Sub-layer (CPS) and Security Sub-layer (SS) which carriers out the authentication secure key exchange and encryption [15]. Figure 3.7.1 shows the MAC layer reference model

Service Specific Convergence Sub layer (CS):

This layer stays on top of the MAC layer architecture. It is a sub-layer which assures the data transmission and takes the data from the upper layer entities. This sub-layer also enables the QoS and the bandwidth allocation. The convergence sub-layer transforms external data through service access point (SAP) into MAC service data unit (SDU) and received by CPS [15].

IEEE 802.16 specifies two types of SSCS for mapping function. First one is ATM Convergence sub-layer which is a logical interface responsible for the ATM. It accepts ATM cells from ATM layer and classifies them. After classification it sends CS PDUs to MAC SAP. The second one is Convergence sub-layer which is a packet based protocol, performs the packet mapping for Internet protocol, Internet protocol version 4(IPv4), Internet protocol version 6(IPv6).

Common Part Sub-layer (CPS)

The CPS represents the kernel of the MAC layer. It stays underneath SSCS and above the SS. It provides major MAC functionalities like bandwidth allocation, system access, establishing connection and maintains the connection-oriented service to SS. The other responsibilities include enabling transmission QoS for service flows and managing the scheduling of data by adding or deleting the connection either statically or dynamically. On the uplink channel it defines three major principles for the transmission such as polling, contention procedures and unsolicited bandwidth permission [16]. The CPS receives data from the various convergence sub-layers through the MAC service access point (SAP)[15].

Security sub-layer (SS):

This sub-layer stays at the bottom of the MAC layer and provides subscribers to the BS with the privacy across the broadband across wireless network [15]. It is most important part of the MAC as it provides authentication, encryption and security key exchange for the system. A secure distribution of keying data from base station to subscriber station is assured by providing authentication. In SS, the addition of a digital certificate strengthens the privacy of data.

3.7.2 MAC Protocol Data Unit (PDU):

The MAC protocol Data unit (PDU) is the basic unit of information exchanged between BS and SS and delivered to the physical layer. It begins with a fixed length generic MAC header. The payload may be followed by the payload, which consists of zero or more sub-headers or zero or more MAC service data units (SDUs). The payload may vary in length and it is represented by variable number of bytes [10]. Figure 3.7.2 illustrates the structure of MAC PDU format

There are two types of MAC headers. The first type of header is Generic MAC header and the other is bandwidth request header for requesting additional bandwidth. The generic MAC header begins each MAC PDU. For the mobile WiMAX, the mobile MAC PDU consists of one generic MAC header with no payload or Cyclic Redundancy Check (CRC). There is a header type (HT) available to distinguish the types of uplink MAC headers. HT is set to zero, if the generic MAC header is followed by payload or CRC. HT is set to one, when no payload is required. For the downlink, HT is always set to zero, since only one MAC header is used. The Generic MAC header is illustrated in Figure 3.7.3

The generic MAC header contains the following fields:

  • EC: The Encryption control (EC) field has got two values, zero indicates that payload is not encrypted and one indicates that it is encrypted.
  • Type: This field indicates the sub-headers and types of special payloads present in the payload.
  • ESF: The extended subheader field (ESF) is available for the DL and UL. If ESF is zero, it means that extended subheader is not present. If the ESF is one, then extended subheader field is present and will immediately follow the generic MAC header.
  • EKS: The encryption key sequence (EKS) tells about the method of encryption that is used to encrypt the payload data.
  • LEN: The length (LEN) field indicates the total length of the MAC PDU in bytes, including the MAC header and the CRC data.
  • CID: This field is unique Connection identifier which is assigned by the BS.
  • HCS: The header check sequence (HCS) is the eight-bit field which is used to detect the errors in the headers.

The MAC header without the payload is applicable only to UL. Payload is optional and it is used to send subheader data. So header format can be used for the signalling services. There are two types of signalling header types available. In the type I header, type field consists of three bits and helpful for designing the bandwidth requests. Type II has got only field in the EC type field. The typical feedback header is the example for this type [15]

3.7.3 Scheduling and QoS Support:

Centralized scheduling at each BS enables the efficient resource allocation in each of the OFDMA frame. Therefore each of the resource allocation in DL and UL in every OFDMA frame is communicated in MAP messages at beginning of each frame. UL bandwidth requests are supported through ranging, piggybacking and polling. The scheduler also supports resource allocation in multiple sub channelization schemes. The MAC layer of the 802.16 standard was developed to support voice, data and video under various busty conditions. The MAC layer manages the radio resources to efficiently support the QoS for the connections established by the BS. Mobile WiMAX standard specifies five scheduling types namely Unsolicited Grant Service (UGS), real-time polling Service (rtPS), non-real-time polling Service (nrtPS), Extended of real-time Polling Service (ertPS) and Best Effort (BE) [9].

Chapter 4: HARQ Mechanism and Overview of TCP Architecture

4.1 Overview of Hybrid Automatic Repeat Request (HARQ) Mechanism:

4.1.1 Introduction:

Hybrid Automatic Repeat Request (HARQ) mechanism is an enhancement to the ARQ mechanism. In the standard ARQ mechanism, only error-detection(ED) information is added for data which has to be transmitted. In HARQ, it adds the existing ED bits as well as the Forward Error Correction (FEC) bits. So HARQ performs better than the ARQ in the poor signal conditions. There are two types of HARQ namely Type I and Type II HARQ mechanisms. The simplest version is Type I HARQ, which adds both the ED bits as well as FEC to the each message before the transmission. Type II HARQ either transmits the ED bits or only the FEC information. HARQ can be used either in selective repeat mode or in stop and wait mode. Since stop and wait mode is simpler, I have used stop and wait mode in my thesis.

4.1.2 Basic Operation:

Stop and wait is an error correction technique where the sender will transmit a block of data and then wait for the acknowledgement from the receiver for transmitting next block of data. If the receiver sends an Acknowledgement (ACK) stating that the data is received correctly without any error. If there is any error, receiver will send the Negative Acknowledgement (NACK) back to the sender. During the normal transmissions, sender will wait for ACK and then transmit the next block of data. In case of longer delays, sender will wait for appreciable time for the response. This state is called the idle state where the sender will not able to send the further data. Figure 4.1.1 illustrates the simple operation of stop and wait technique.

The sender sends the first block of data and receives the ACK from the receiver. An ACK field has the next expected packet sequence number. Sender keeps on sending frames till it receives NACK from the receiver. When a frame is damaged or corrupted, receiver will send the NACK. Since the third block of data is damaged, it gets back the NACK and retransmits the frames starting from the rejected frame.

Figure 4.1.2 Operation of HARQ mechanism

Fig. 1 illustrates a single TCP DATA packet delivered from the content provider located in the fixed part of the network to the Mobile WiMAX Station (MS). The Base Station (BS), after reception of a TCP data packet from the fixed network (File Server), forwards it over the radio link to the appropriate MS. This downlink transmission includes also heavy physical and link layer overhead. In case of successful reception, the MS sends an HARQ ACK response to the BS and forwards the received TCP data packet up to the TCP layer of the protocol stack. Consequently, the TCP layer generates a TCP ACK. This one represents ordinary payload for the WiMAX link layer: before it can be transmitted on the wireless link, uplink resources should be requested and a corresponding assignment grant should be received at the MAC layer. Moreover, TCP ACK transmission requires an HARQ ACK from the BS.

4.3 Overview of Transmission Control Protocol (TCP):

4.3.1 Introduction:

Transmission control Protocol (TCP) is the largest Internet Protocol which carries 90% of the traffic in todays heterogeneous wireless networks.TCP is widely used as a connection oriented protocol to provide reliable packet delivery over unreliable links[17][18]. If TCP directly runs over the wireless networks, then there will be packet loss due to the channel errors and it will be treated as loss due to the congestion in the network. So the congestion control will be invoked, so data transmission rate will be reduced and causes degradation in the throughput [19]. In my thesis, I have used TCP because it supports retransmission which is needed for HARQ mechanism to recover the packet losses.

4.3.2 Types of TCP flavors:

Various TCP flavors are used to reduce the congestion in the network. There are four major types of flavors used in the TCP.

4.3.2.1 Tahoe:

TCP Tahoe was the first algorithm to use three transmission phases such as slow start, congestion avoidance and fast retransmit. Slow start and congestion avoidance helps to increase the window size. Fast retransmit is used to detect congestion [20]. Since it does not suitable for recovery in packet losses, this flavor is not suitable for my thesis.

4.3.2.2 TCP Reno:

TCP Reno is the modified version of TCP Tahoe. It employs four transmission phases namely slow start, congestion control, fast retransmit and fast recovery. If there is any packet loss due to congestion, either sender receives three duplicate acknowledgements or retransmission timeout (RTO) timer expires. Upon three duplicate acknowledgements, the sender activates the fast retransmit and recovery algorithms. Since TCP Reno uses the fast recovery algorithm only for a single packet loss [21]. This version degrades the performance in case of multiple packet losses, so it is not possible to use this flavor in my thesis.

4.3.2.3 TCP New Reno:

TCP New Reno is a modification of TCP Reno. It improves the retransmission process during the fast recovery phase. Unlike TCP Reno, this version can detect multiple packet loss and will not exit until all unacknowledged segments of fast recovery are acknowledged. It retransmits the next segment only it gets the partial acknowledgements. Partial acknowledgements are the one which does not acknowledge the outstanding packets upon the fast recovery [22]. Since the receiver does not transmit the missing data segments, this version/flavor is not suitable for my thesis.

4.3.2.4 Selective Acknowledgement (SACK):

SACK is used to acknowledge the out-of order segments selectively rather than cumulatively acknowledging the last correctly in order received segment. The advantage of having SACK is that the receiver acknowledges the packets which are received out of order and the sender retransmits only the missing segments instead of sending all unacknowledged segments. TCP with SACK helps to improve the performance in case of multiple packet losses. During the fast recovery phase, SACK maintains a variable called pipe to estimate the number of outstanding packets [23]. Since this version/flavor saves time and does the fast recovery only for the packets that are received out of order, I have used this flavor in my thesis.

4.4 Protocol Architecture:

The TCP architecture is made up of four layers which work together between themselves. The five layers from high to low are:

  1. Application layer
  2. Transport layer
  3. Network layer
  4. Link layer

The Transport layer is very important layer which has the capabilities like error control, segmentation, flow control, congestion control. The transport layer make sure that the data arrives in order, discards the duplicate data and makes sure that the discarded packets are resent again. Each segment created by TCP has a sequencing number of re-ordering after receipt, an acknowledgement ID number, source address, destination address and check sum for error correction, window, offset and urgent pointer.

TCP uses a sequence number to identify each byte of data. The sequence number is helpful to identify the order in which the packets are sent so that it can be reconstructed in order regardless of packet loss, fragmentation. TCP uses the end-to end flow control to avoid the sender from sending data too fast to the receiver. The important aspect of TCP is congestion control. The senders employ retransmission time out (RTO) and it based on round-trip time (RTT) for effective retransmissions. The Maximum segment size (MSS), normally specified in bytes, is the largest amount of data sent by TCP. The value of MSS should be low as possible to avoid excessive retransmissions. TCP also uses retransmission timer, time-waited timer and persistence timer for providing reliability for the data. User data protocol(UDP) is suitable for real time applications like video conferencing and voice, but it does not corrects the error and not reliable for the data. So i have used TCP in my thesis which detects the error and also reliable for the security of the data. All retransmissions are done in transport layer because it detects the error and does the FEC and passes the data to the link layer.

Chapter 5: Method of the project and performance analysis

5.1 Introduction

Simulation can be defined as the process of creating abstract representation model for an existing model. There is some simulation software available in the market like OPNET, NS-2, Commsim, etc. These valuable tools provide a procedure that can be evaluated without working on real-time scenario. Many company professionals use these tools to analyze their performances as these are cost effective, easily available in the market and easy to install.

The simulation tool used in my thesis is OPNET modeler 15.0, because the simulation results produced by OPNET are highly reliable and efficient. It is also very easy to evaluate the performance of the existing model and can be easily replicated in the real time scenario. Also the software does not require any separate programming like other simulators. So it is the suitable software for calculating the performance evaluation for a WiMAX network.

5.2 Simulation Methodology:

In my thesis, I have used OPNET 15.0 to simulate performances in dynamic virtual network environments such as protocols, architectures. The method used in my thesis is by creating a basic network, naming the empty scenario and then setting up for a model family. The basic architecture have at least one Base station (BS), one or Subscriber Stations (SS) which are connected to the core network. All of these are available in the object palette. The node models mainly used in the WiMAX network architecture for my thesis are

  • Application Definition
  • Profile Definition
  • WiMAX configuration
  • Base station
  • Subscriber station
  • WiMAX server

Application Definition:

The application definition has to be configured with its attributes to evaluate the performance of the proposed network. The applications used in the application definition are VoIP and video conferencing. In my thesis, i have chosen TCP protocol and used the digital bandwidth of 10 MHz and analog bandwidth of 11 Mb/s.The application definition is represented as below:

Profile Definition:

The profile definition attributes are configured accordingly to the application defined in the application definition. I have used the profile for the voice which has the parameters like offset start time, start time duration and end time duration, repeatability and operation mode which are fixed according to the simulation. The profile definition is represented as follows:

WiMAX configuration:

The WIMAX configuration module is main node which has the exact attribute to configure the network. It has the attributes like efficiency modes, MAC service class definitions, OFDM PHY profiles, and channel coding. I have used physical layer efficiency mode which supports HARQ mechanism.

Base station:

The base station has to be configured accordingly so that we can perform the evaluation of HARQ mechanism in it. The base station has the parameters like antenna gain, maximum and minimum power density, classifier definitions and maximum transmission power.

Subscriber Station:

The WiMAX MAC uses scheduling algorithm for which the subscriber station will compete once for the initial entry into the network. After assigning the network entry, the subscriber station will be allocated an access slot. The subscriber station has various parameters like service class name, modulation and coding, uplink service flows and downlink service flows. In my thesis, I have used QPSK modulation and enabled HARQ attribute. The Subscriber station configuration is illustrated in Figure 5.2.4

WiMAX Server:

5.3 Traffic Model:

To analyse the QoS in real time scenarios, i have implemented VoIP in real time traffic and it works in both commercial and domestic services. The traffic specification for voice application is given in table 5.3.1. I have chosen G.711 voice codec which uses Pulse code modulation (PCM) samples for the signals of voice frequencies, sampled at a rate of 8000 samples/second resulting in 64 Kbit/s. It reduces the usage of bandwidth by 50%. There are seven types of service available. The default service type is Best Effort type. I have chosen the Interactive Voice which has the highest priority than the best effort. The traffic mix can be selected as pure discrete or pure interactive voice. I have used traffic mix for the VoIP as 25% and also only one voice frame per packet is chosen to avoid the network congestion

5.4 Performance Metrics

The Performance Evaluation for the HARQ mechanism in mobile WiMAX is evaluated using the following metrics:

Mean Opinion Score (MOS):

Mean Opinion Score (MOS) is qualitative measurement of voice quality on a conversation. The value ranges from 0(pure noise) to 5(perfect fidelity). MOS value is based on the codec choice, packet loss, packet delay variation, packet end to end delay and jitter. Any value over 3.5 is considered good for the voice.

Jitter:

Jitter is the variation in the time between the packets transmitted from source to destination, caused by the congestion, timing drift. Jitter is an important parameter for the voice and the value of jitter should be as low as possible.

Delay:

Delay occurs when packets of data takes more time than expected time to reach their destination. Higher delay causes more disruption in the voice quality, so lower delay is very important for the real time traffic to improve the throughput.

Packet End-End delay:

The time taken for a packet to be transmitted across the network from source to destination.

Packet delay variation:

Packet delay variation is the difference in the end to end delay between the selected packets.

Throughput:

The throughput is defined as the average rate of successful message delivery over a communication channel. The throughput is usually measured in bits per second or data packets per second. Higher throughput implicates the improvement of QoS for the network.

Transmission Rate:

The average transmission rate is the rate at which the information is processed across the network. It is calculated as the number of bytes or packets against the time interval.

5.5 Performance analysis and Simulation results:

This section provides the results and analysis of my thesis obtained by simulating the proposed WiMAX network. It also provides the graphical representation of the metrics evaluated by performing simulation for the three scenarios. There are three scenarios used in my thesis. In the first scenario, the simulation results are obtained without enabling HARQ mechanism. In the scenario, the features for HARQ mechanism is enabled and results are compared with the first scenario. In the last section of the simulation, modifications are done by adding ARQ mechanism with HARQ mechanism. Finally the modified scenario of using both ARQ and HARQ shows that it can perform better than the existing HARQ mechanism.

5.5.1 Simulation Setup:

Before conducting the three scenarios, basic simulation parameters have to be configured using application and profile definition. In the application definition, the parameters for VoIP such as packet size, type of encoder, traffic mix, type of service class, etc has to be configured. In the profile definition, the application start time and end time has to be defined for the VoIP application. The base station has to be configured with MAC classifier definitions, Maximum power density and minimum power density has to be configured. The subscriber station has to be configured with the HARQ parameters such as number of retransmissions, number of channels available for uplink and downlink. For the first scenario, HARQ parameters are disabled in the Subscriber stations and in the WIMAX server. In the second scenario, the HARQ parameters are enabled and the results are compared with the first scenario. In the final scenario, the modified version of HARQ mechanism yields better result than the existing HARQ mechanism with the addition of ARQ mechanism with the HARQ mechanism.

Scenario

[1] http://www.tech-ware-tips-startup-internet-business.com/Wireless-LAN-uses.html

[2] http://www.tutorial-reports.com/wireless/usage.php

[3] www.ebooksportal.com, WiMAX Operators Manual Building 802.16 Wireless Networks 2nd Edition

[4] http://en.kioskea.net/contents/wireless/wman.php3

[5] www.knownfree.com, WiMAX Taking Wireless to the MAX.

[6] http://www.networkworld.com/news/tech/2001/0903tech.html

[7] http://en.kioskea.net/contents/wimax/wimax-intro.php3

[8] Mohammad Azizul Hasan, Performance Evaluation of WiMAX/IEEE 802.16 OFDM Physical Layer, Espoo, Finland, June 2007

[9] Kwang- cheng chen, J. Roberto B.de Marca, Mobile WiMAX, John Wiley & sons , Ltd,2008

[10] Syed Ahson , Mohammad Ilyas WiMAX standards and Security

[11] Loutfi Nuaymi , WiMAX: technology for broadband wireless access

[12] Pandharipande, A.; Principles of OFDM, IEEE Potentials, Volume 21, Issue 2, Apr-May 2002.

[13] Eklund, Carl, et.al. WirelessMAN, Inside the IEEE802.16 Standard for Wireless Metropolitan Networks, IEEE Press, 2006.

[14] Jeffrey G. Andrew, Arunabha Ghosh, Rias Muhamed : Fundamentals of WiMAX: Understanding Broadband Wireless Networking

[15] Yan Zhang, Hsiao-Hwa Chen, Mobile WiMAX: toward broadband wireless metropolitan area networks

[16] Convergence Sublayer and Common Part Sublayer, http://www.cs.tut.fi/~83180/83180_05_S8c.pdf

[17] J. Postel, Transmission control protocol, RFC 793, Sept. 1981.

[18] A. Gurtov and S. Floyd, Modeling wireless links for transport protocols, ACM SIGCOMM Comput. Commun. Rev., vol. 34, no. 2, pp. 8596, Apr. 2004.

[19] Y. Tian, K. Xu, and N. Ansari, TCP in wireless environments: problems and solutions, IEEE Commun. Mag., vol 43, no. 3, pp. S27S32, Mar. 2005.

[20] V. Jacobson, Congestion avoidance and control, in Proc. ACM SIGCOMM, Symp. on Commun. Archit. And Protocols, Stanford, CA, USA, Aug. 1988, pp. 314329.

[21] S. Floyd and K. Fall, Simulation based comparisons of Tahoe, Reno, and SACK TCP, ACM Computer Communication Review, vol. 26, no. 3, pp. 521, July 1996.

[22] S. Floyd and T. Henderson, The New Reno modification to TCPs fast recovery algorithm, RFC 582, Apr. 1999.

[23] M. Mathis, J. Mahdavi, S. Floyd, and A. Romanov, TCP selective acknowledgement options, RFC 2018, Oct. 1996.

Please be aware that the free essay that you were just reading was not written by us. This essay, and all of the others available to view on the website, were provided to us by students in exchange for services that we offer. This relationship helps our students to get an even better deal while also contributing to the biggest free essay resource in the UK!