Hypergraph-based resource allocation in wireless mesh networks

The success of wireless mesh networks (WMNs) is highly contingent on effective radio resource management. In conventional wireless networks with heterogeneous traffic, system throughput, quality of service (QoS), fairness and complexity are usually common performance metrics. Orthogonal Frequency Division Multiplexing Access (OFDMA) is the most known modulation technique in wireless networks which are largely dependent on the mechanism of resource allocation. Limited spectrum resources necessitate that these resources be optimally employed in such a way as to satisfy the requirements of WMN users. Therefore, the task of resource allocation in this technology should be formulated as a task of allocating resource blocks (RBs) to the network stations according to the required traffic demands and QoS parameters and should also take into account the interference among them. By considering these features in mathematical models using a hypergraph and Koenig graphs, we can obtain an accurate description of the network. This enables us to fully describe all possible configurations of a mesh network and its individual elements, and takes into account the limitations of spectrum resources and interference between the network stations. The proposed mathematical model was aimed at improving the performance of a mesh network as a whole by balancing the number of resource blocks allocated to individual stations. Thus, our algorithm was developed such that the throughput was maximised as long as the QoS was guaranteed with low complexity. Finally, a hypergraph-based resource allocation in a wireless mesh network with a balanced technique (BA-HBRA) was developed for OFDMA downlink RBs allocation to further enhance the system performance by satisfying the traffic demands with the least number of RBs and simultaneously ensuring the QoS. Algorithms, OFDMA-based channel-width adaptation in a wireless mesh network (OBCWA) and QoS-aware channel-width adaptation (QACWA), were compared. The results of simulations showed that the throughput of BA-HBRA outperformed OBCWA by 13%, 12%, 8%, 8%, 7% and 10%, respectively, in six cases with different numbers of RBs. In terms of QoS provision with low traffic demands, our algorithm achieved the same results as QACWA and it could guarantee the QoS. When the traffic demands increased with different services, BA-HBRA throughput achieved about 12%-25% more than QACWA, and was close to satisfying the traffic demands in most cases. When the traffic was relatively high, QACWA could not satisfy the traffic demands. Finally, in terms of complexity, we found that the processing level of our algorithm was the same as QACWA. Although they were at the same level of processing, BA-HBRA achieved a remarkable improvement over QACWA in terms of time consumption, where BAHBRA took about 82%, 66% and 40% less time than QACWA did. At the same time, it also outperformed OBCWA with about 87%, 78% and 63% for three sets of nodes (10, 20 and 30, respectively). From the above simulations, we can conclude that our proposed algorithm is more suitable for WMNs, especially those with limited-resource scenarios.

Preview Text FK 2018 60 IR.pdf Download (1MB) | Preview

Actions (login required)

View Item