OPTICAL NETWORK DEMAND BUNDLING USING DS3-FORMING
The survivability of communication networks in the physical layer topology has been discussed from Fiber Span Layout Demand Distribution and Fiber Network User Service Survivability (FNUSS) in the previous chapters. This chapter describes the Optical Network Demand Bundling using DS3-Forming [ONDB] in the physical layer Module. This model comprises of the Demand Distribution in terms of parcel list in the physical layer.
From the evaluation of Digital Level characteristics, Vande.V et.al. discussed the methods of survivability, which can be provided in single fiber network architecture as shown in the Fig. 4.1,4.2,4.3 and 4.4. However, their architecture supports low capacity network cross connectivity of fiber optic facility links. They estimated the Digital Level formations individually to a limited number of Demands of fiber optic facility links and 20% throughput is obtained.
The present discussions and simulations in Optical Network Demand Bundling [ONDB] supports all the Digital Levels. This type of architecture is very much essential from the survivability of Digital Level Formations point of view. This architecture provides the implementation of facility hubbing and facility hierarchy methodologies and includes Diversity technique and is extended to Demand Bundling Technique, a new algorithm.
This process is proposed and implemented successfully to estimate the parameters like Nodal Connectivity Requirement (NCR) and also the Network Survivable Architecture (NSA) at the DS1 and DS3 levels. It is further extended to Optical Networks System (ONS) in a single Period demand bundling and also multiperiod demand bundling. In this process routing takes place in direct and indirect path. Direct path means a DS1 that does not rearrange signals at an Intermediate office. Indirect path consists of two or more DS3s formed by rearranging signals at an intermediate hub location. This method is affordable and also provides the network planners with flexible demand requirement.
4.2 FIBER NETWORK USING DS3-FORMING
The present work represents the evaluation of the Demand Bundling, in the fiber network, both directly and indirectly with respect to different demands and also Facility Hubbing with Diversity Protection is as shown in the Figs. 4.5, 4.6 and the program is given in appendix C.
The demand bundling procedure and its algorithm combines point-to-point links into appropriate DS3-level demands, Flowchart are as shown in the Fig. 4.7, Fig.4.8 and the program is given in Appendix - C. This process explains that each DS1 or DS0 demand pair would route over a unique DS3, thus resulting in low filled DS3 systems. The DS3 (or STS-1) signal level is commonly used as input to fiber systems in today's interoffice fiber networks .
The Demand Bundling Algorithm consists of two sub algorithms. First, The single period bundling algorithm that uses a demand bundling concept in hierarchical networks which assigns traffic to circuits. The second algorithm is a more sophisticated, multi-period demand bundling algorithm that is performed in two phases. Phase1 solves a single period static model, which relaxes capacity constraints in order to retain the demands with the most effective routes. Phase2 performs a more detailed, multi-period optimization on each routing path chosen in phase1 .
The resulting trunk capacity requirements are used for the bundling algorithm process. The demand bundling algorithm, which routes circuits over DS3s, uses a similar hierarchical approach to evaluate the economics of routing circuits directly between two CO's or using an intermediate hub DCS to further aggregate the demand . The DCS discussed here is a DCS 3/1 that terminates DS3s and cross connects DS1s. The Routing mechanism implies the following points in the DS3 Forming levels:
* CO-to-home hub DS3, which originates from local CO demand and terminates on the home hub DCS of that CO.
* CO-to-foreign hub DS3, which originate in a CO and terminates on a hub DCS other than the home hub of that CO.
* Hub-to-hub DS3s, which terminates on a hub DCS 3/1 at both hubs.
It builds direct DS3s for each CO, and then processes the next higher facility hierarchy. Further it completes its process by a path from the highest gateway to the home hub and back to the CO.
Once the number of DS1's is obtained the formation of Co-to-CO, DS3s is calculated such that the line rate of the number of DS1's should not exceed the line rate of DS3 for Direct DS3 forming. Hence ONDB connectivity performance is measured in terms of Direct/Indirect DS1/DS3 to the demands in a given network.
4.3 NUMERICAL RESULTS
Numerical Results have been evaluated for a complex network with multi connectivity in DS3 Forming. The input connectivity values for DS3\DS1 for Direct and Indirect DS3/DS1 cases are as shown in the table 4.1. The demands assumption is at random. The simulation output results such as connectivity pattern values for Direct/Indirect DS3/DS1 are presented in table 4.2.
In the previous work Vande. V. et.al. Group worked  for single fiber network architecture and achieved 25% end-to end demand bundling strategy. By using the ONDB DS3-Forming, bundling technique connectivity performance achieved is 85% as compared to the work reported earlier.
In this work, a new algorithm is proposed i.e. Demand Bundling, which extends point-to-point connections to appropriate DS3-level demands. It results in two sub proposals, namely a single demand bundling algorithm that uses a concept in designing the hierarchical networks and the second one is a more sophisticated multi-period demand bundling algorithm.
The Multiplexing is a network model that addresses the worst case demands where all demands for a multi-period problem are summed into different parcel lists. Using this concept the model deals with the routing of DS3 systems over fiber spans, which are fixed during all periods. Increase in demand on any fiber span in any single period should not be greater than the maximum line rate. The computation levels in DS-3 forming extend to a dynamic programming model, to achieve the optimum level network survivability.
Survivability parameters of Optical Network Demand Bundling using DS3 Forming in the Physical Layer Module has been analyzed in this chapter, and the procedure is extended to compute the integrated parameters of the optical networks and is discussed in the next chapter.
Tables 4.1 INPUT TABLES
Connectivity pattern and values for Direct DS1s, DS3s
DIRECT DS1s & DS3s
(1, 2) à 13,15
(1, 3) à 17,21
(2, 3) à 18,25
(3, 7) à 16,30
Connectivity pattern and values for Indirect DS1s, DS3s
DIRECT DS1s & DS3s
(1, 4) à 6 (4, 1) à 6
(1, 5) à 9 (5, 1) à 10
(1, 6) à 9 (6, 1) à 10
(1, 7) à 9 (7, 1) à 9
(1, 8) à 6 (8, 1) à 7
(2, 8) à 6 (8, 2) à 5
(3, 5) à 7 (5, 3) à 7
(4, 8) à 2 (8, 4) à 3
(5, 7) à 8 (7, 5) à 8
(6, 8) à 4 (8, 6) à 4
Table 4.2 DEMAND BUNDLING OUTPUT TABLE
DIRECT DS1s & DS3s
(Sum of DS1&DS3 in different parcel lists)
(1, 2) à 26
(1, 3) à 34
(1, 4) à 12
(1, 5) à 19
(1, 6) à 19
(1, 7) à 18
(1, 8) à 13
(2, 3) à 30
(2, 6) à 11
(3, 5) à 14
(3, 7) à 25
(4, 8) à 5
(5, 7) à 16
(6, 8) à 9