Research On Sensing Based Partial Overlapping Channel Assignment In The Public Access Network

As the wireless network continues wave into the people’s daily life, public access network dominated with its various advantages. As the typical representative of the public access network, WLAN networks and wireless Mesh networks have gained unanimous approval, and widely deployed in practice, for both of them predominant in architecture and performance.But we have to face the problem that usually the channel resource is not utilized sufficiently in a public access network. Most of the wireless networks are using the orthogonal channels for communication, so the presence of the partially overlapping channel resources completely wasted. Although researchers have proposed methods for ultilizing the overlapping channel resources, but most of them are based on location information. These methods are based on two very strong assumptions, one known as the position of the nodes in networks are clearly known, the second is the signal propagation in a free space model. As our knowledge, in the actual networks, these two assumptions are very difficult to meet. Because in actual network deployment, not only the nodes involved in complex three-dimensional space, the location information will become unavailable or too difficult to use, but also between the buildings the network nodes deployed in, the signal propagation is affected by diffraction, reflection, scattering and other diverse radio characteristics. Therefore, the methods designed for using overlapping channels efficiently is imperative. This article designed the sensing based partial overlapping channel assignme nt basing on the consideration showed above.The sensing based partial overlapping channel assignment method overcomes the shortcomes introduced by the two assumptions in position information based method. In the sensing based partial overlapping channel assignment method, through sensing the surrounding nodes or links interference energy, the working channel and other related informations, the node can firstly convert the relative position of himself in the network into a local weighted conflict graph. After that, solving such a local conflict graph, that is a constraint satisfaction problem, thus, assign the most appropriate channel for the currently referenced node. At the same time, a loosing processing for the constraint satisfaction problem solving whe n the constraint can not satisfied introduced. Meanwhile, a penalty value is used to estimate the interference instroduced by the loosing process.In this article we proposed CI_SBPOCA, a partial overlapping channel assgnment algorithm based on channel information sensing for WLAN, which select a minimum interference energy channel as the working channel. This method is the most simple, but working in a partial greedy way, so the performance is not so good; Additional, we proposed NI_SBPOCA, an algorithm based on node state information sensing, which adopted the heuristic methods to maximize the selection for following nodes and optimize the channel selection process. After that we proposed an integrated sensing based channel assgnment algorithm for wireless mesh network.At last, the proposed algorithms are experimentally verified using the NS3 network simulation tools. We find that the sensing based partial overlapping channel assignment method is greatly improved in performance compared to the orthogonal channel allocation method. Meanwhile, comparing to location-based partial overlapping channel assignment method, sensing based method can greatly improving the availability in the actual deployment while ensuring the performance.