Adopting dynamic operators in a genetic algorithm

Journal of Computer and System Sciences

Ikeda

et al. 2015

Wireless Mesh Networks (WMNs) are attracting a lot of attention from wireless network researchers. Node placement problems have been investigated for a long time in the optimization field due to numerous applications in location science. In our previous work, we evaluated WMN-GA system which is based on Genetic Algorithms (GAs) to find an optimal location assignment for mesh routers. In this paper, we evaluate the performance of four different distributions of mesh clients considering Packet Delivery Ratio (PDR), throughput and delay metrics. For simulations, we used ns-3 and Hybrid Wireless Mesh Protocol (HWMP). We compare the performance of the optimized network component for Normal, Uniform, Exponential and Weibull distributions of mesh clients by sending multiple Constant Bit Rate (CBR) flows in the network. The simulation results show that for Normal distribution, all mesh routers are concentrated in the center of grid and the communication becomes easy. On the other hand, for Uniform distribution the mesh routers are more scattered, the creating of links is more difficult and the communication can be done only with multiple hops. The increase of number of connections, increases the traffic data rate and the network load causing a decrease of the average PDR and throughput. When the number of mesh routers is increased the number of hops increases and PDR decreases. Also, for Weibull distribution, when the number of mesh routers is increased the number of hops increases and PDR decreases. When there are 35 mesh routers in the network with 10 connections, the PDR is higher compared with other cases. For Exponential distribution, the throughput is higher for small number of mesh routers. On the other hand, for Weibull distribution, when there are 20 connections, with the increase of the number of mesh routers the throughput is decreased much more than the case of Exponential distribution. The delay is almost the same for Exponential and Weibull distributions.

Section: Selection Operatorsmentioning

confidence: 99%

Node placement for wireless mesh networks: Analysis of WMN-GA system simulation results for different parameters and distributions

Barolli

Journal of Computer and System Sciences

Ikeda

et al. 2015

“…Selection Operators: In the evolutionary computing literature we can find a variety of selection operators, which are in charge of selecting individuals for the pool mate [26]. The operators considered in this work are those based on Implicit Fitness Re-mapping technique.…”

Section: Algorithm 1 Genetic Algorithm Templatementioning

confidence: 99%

Node Placement in WMNs Using WMN-GA System Considering Uniform and Normal Distribution of Mesh Clients

2014 Eighth International Conference on Complex, Intelligent and Software Intensive Systems

Sakamoto

Spaho

et al. 2014

In this paper, we evaluate the performance of WMN-GA system for different settings of number of mesh clients. The population size is considered 64 and the number of generation 400. We considered Uniform and Normal distributions of mesh clients. For evaluation, we consider average size of giant component (Avg. SGC) and average number of covered mesh clients (Avg. NCMC) metrics. The simulation results shows that for both distributions, with the increase of the number of mesh routers, the number of covered mesh clients is increased. With the increase of the number of mesh clients, the percentage of Avg. NCMC is increased. Comparing the simulation results for both Uniform and Normal distributions, for number of mesh routers 16 to 64 the performance of Normal distribution is better compared with Uniform distribution, but for number of mesh routers 128 for both distributions the performance is very good.

“…In the evolutionary computing literature we can find a variety of selection operators, which are in charge of selecting individuals for the pool mate [15]. The operators considered in this work are those based on Implicit Fitness Re-mapping technique.…”

Section: A Selection Operatorsmentioning

confidence: 99%

A Comparison Study of GA and HC for Mesh Router Node Placement in Wireless Mesh Networks

2013 16th International Conference on Network-Based Information Systems

Chang

Sakamoto

et al. 2013

Wireless Mesh Networks (WMNs) are important networking infrastructure for providing cost-efficient broadband wireless connectivity. Mesh router node placement is important to achieve network connectivity and coverage in such networks. In this paper, we compare the performance of Genetic Algorithm (GA) and Hill Climbing (HC) for mesh router node placement in WMNs. We consider Exponential and Weibull distributions of mesh clients and grid size 16x16, 32x32 and 64x64. As evaluation metric we used size of giant component. Simulation results show that GA performs better than HC.