Neural Network and Genetic Algorithms for Topology Optimization of the CCS7 Network

Abstract

The topology optimization of CCS7 network was formulated in Xin and Xu (1998); the A/B plane partition of HSTPs in CCS7 network is discussed here in detail. The problem is proved to be NP-complete, so neural network and genetic algorithms are applied. With test networks generated randomly, the computing results show that the genetic algorithm is quite robust. Topology of CCS7 networkCommon Channel Signaling (CCS) is one of the signaling methods by which a separate (from service circuits) common channel conveys signaling information for multiple service circuits. CCS network is the key support network for modern telecommunication networks. CCS No.7 (CCS7) and the CCS7 network are the standard recommendations of CCITT (CCITT Blue Book, 1988).The CCS7 network consists of signaling points (SPs), signal transfer points (STPs) and signaling links (SLs). SP is a center as an origin and destination of signaling messages. STP transfers the signaling messages from one link to another. SL is a digital transmitting circuit, which connects SP to STP, and STP to STP.The technical standard of Chinese CCS7 network, referring to the recommendations of CCITT, provides that the Chinese CCS7 network employs a three-level hierarchy system with A/B plane in first-level, and adopts the quasi-associated mode of signaling:(1) STPs at the first level (HSTPs) are partitioned into A and B planes; HSTPs in each plane are fully connected, but only the paired HSTPs between the two planes are connected. A pair of HSTPs is home STPs in a first-level signaling region.(2) Each STP at the second level (LSTP) is connected to its two home HSTPs respectively.

Ripple-Spreading Network Model Optimization by Genetic Algorithm

“…As large-scale parallel stochastic search and optimization algorithms, genetic algorithms (GAs), if properly designed, have the capability of producing high quality solutions to NP-hard problems in an acceptable period of time [19]. Actually, GAs have already been used to optimize some network structures, for example, the topology optimization of CCS7 network [20], MPLS network [21] and airline route networks [22]. However, in such studies on network topology optimization, search is carried out by directly adding new links or removing existing links between nodes.…”

Deterministic ripple-spreading model for complex networks

“…As is well known, the optimization of network topology is a NP-hard (non-deterministic polynomial-time hard) problem. Population-based algorithms, such as genetic algorithms (GAs), have the potential to resolve this problem [14][15][16], but the widely used data structures, such as adjacency matrix and list of edges, to record network topologies may usually jeopardize the scalability and practicability of population-based algorithms. Basically, the adjacency matrix is memory expensive and therefore can be used hardly by population-based algorithms to optimize large-scale networks.…”

Application of Complex Network Theory and Genetic Algorithm in Airline Route Networks

“…These hub-and-spoke networks allow airlines to benefit from cost and demand advantages (4). In contrast, some new or recently started airline companies continued operating point-to-point networks on a low-cost, quality network topologies (15)(16)(17). The design of evolutionary operators (i.e., mutation and crossover) is particularly crucial in the successful implementation of GAs for such problems.…”