The Joint Spectrum And Power Resource Allocation In Wireless Cognitive Networks

In recent years, the variety ofthe wireless client applications and the popularity ofportable communication devices have brought a dramatic shortage of the spectrumresources.Cognitive radio technology can detect spectrum environment and can makethe cognitive users opportunistic access the idle spectrum, thusit is considered to be oneof the key technologies to solve the spectrum shortage.Resource allocation in Cognitive Wireless Network, especially the spectrumallocation and power allocation are the main factors affecting the system performance.The traditional way focused on spectrum allocation or power allocation independently,but the allocation scheme combining the both is relatively scarce. This kind of jointallocation scheme is recognized as a NP-Hard problem, so the existing joint allocationstrategiesare almost all sub-optimal solutions.This article hasproposed three related joint allocation strategies that canmeetthedifferent needs. Firstly, this paper introduced the convex optimization model, geometricprogramming and dual decomposition theory. Through the model derivation, based onthe classical system model we finally designed the joint allocation algorithm-CJAalgorithm.The algorithm is divided into two steps: the first step is dual decompositionheuristic channel allocation; the second step is convex optimization adjusting powerallocation. Moreover, our article gavethe simulation results and comparative analysisbetween CJA algorithm and the theory optimization algorithm.Secondly, thispaper also concerned some other features of cognitive wirelessnetworks and proposed another joint resource allocation model that is based on users’satisfaction, thisnew model’s optimization objective shifted from traditionalmaximizingthe network throughput capacity to maximizingthe users’ satisfaction, and gavethedefinition of users’ satisfaction that considering both the historical allocation andQoS requirements. And then we designed the joint allocation algorithm-SF-CJAalgorithm, this algorithm needs two important sub-algorithms: the feedback channelselection algorithm and the weight factor update algorithm. In addition, in order toconduct a comprehensive analysis of the simulation, we designed8kinds of test casesto verify the accuracy and feasibility of the SF-CJA algorithm.Finally, based on the country’s first cognitive radio test system, the existing simplecentral allocation mechanism has been improved to be a multi-channel multi-interfacejoint allocation algorithm-M-CJA algorithm, this algorithm can select three channelsfor the whole network, at the same time, it can optimize the transmission power forevery users’three interfaces. The simulation results showed that the M-CJA algorithm ismore flexible than the original strategy, but due to time constraints, the M-CJAalgorithm has not yet join the Linux kernel protocol stack.