Adaptive and optimized radio resource allocation algorithms for OFDMA based networks

Wireless networks based on 4G standards like WiMAX and LTE have embraced OFDMA as the downlink technology due to its capabilities and flexibilities. However, for highest possible network performance and bandwidth efficiency, the radio resources must be distributed very carefully. When assigning subcarriers and power to the connected subscriber stations in a network, a user equipment should be allocated with the subcarriers according to their channel conditions, maintaining appropriate level of service agreement and other necessary conditions such as priority, delay etc. This allows the network service providers to accommodate more users to increase their revenue using the same finite spectrum and yet maintain the expected QoS. The higher the number of suitable subcarriers that can be judiciously assigned to a SS, the more the SS will gain by way of a higher performance contributing to total higher data rate efficiency. Moreover, when a BS is overloaded to try to accommodate a high number of users to provide them with different kinds of services, the performance of the network such as throughput and delay will naturally degrade. Also, users with better channels may continuously enjoy highest amount of radio resources, starving others with poorer channels or preventing new users from entering the network. In OFDMA, subchannels can be shared optimally and subcarriers reallocated cleverly to bring a system-wide fair performance such that the best of the disadvantaged users can still survive while the higher priority users are not adversely affected. Thus, in this thesis, an adaptive and optimized radio resource allocation algorithm (AORAA) with cross-layer mechanism is proposed. It is shown that AORAA enhances the performance of an OFDMA based network in terms of throughput, delay, jitter and fairness. AORAA contains an adaptive and optimized subcarrier allocation algorithm which uses graph theoretic techniques to do the best probable matching of subcarrier and users’ channel information. Also it considers the problem of making the total allocation cycle efficient to be calculated and thus separates the power allocation from doing it simultaneously with optimized subcarrier allocation. It considers different types of user requirements. As users can be categorized into two types, which are service sensitive such as specific delay and rate requirement, and the other is for those who do not have specific service delay or throughput requirements, AORAA considers this and contains the solutions with required formula and algorithm to implement. By this way the different users get their expected fairness and better network experience which is measured and shown in terms of throughput, delay and jitter. Furthermore, a network may heavily loaded compared to its existing resources. The system (AORAA) then should be able to minimize the adverse effects and provide undistorted network service as much as possible. This thesis proposes a threshold based resource allocation that helps to schedule and grant resources intelligently to solve the problem alluded to above resulting in better overall system performance, enhancing the network experience for the users. All the solutions provided in this thesis have been described in it and then presents with adequate simulation results and analysis. It is shown that AORAA outperforms existing referenced works by up to 33% for throughput and 23% for delay on average with better fairness.

Preview Text FK 2015 93 edited.pdf Download (1MB) | Preview

Actions (login required)

View Item