Research On Spectrum Aggregation Algorithm Of Cognitive TD-LTE System

With the transition from the analogue broadcast television system to the digital one, a large number of TV channels are freed up, which are called TVWS (TV White Space). In2010, FCC (Federal Communications Commission) released the final rule to enable unlicensed broadband wireless devices to utilize TVWS. LTE (Long Term Evolution) requires more spectrum resources in order to support more numerous services. But spectrum resources are non-regenerative. LTE as a secondary system can utilize TVWS resources by adopting Cognitive Radio (CR) technique while not bringing about interference to TV system. As in TD-LTE the uplink and downlink share the same spectrum band, TD-LTE is preferable to limited TVWS resources. The available spectrum resources of TVWS are discrete in frequency domain, so it is required to adopt Spectrum Aggregation (SA) technique to aggregate several discontiguous spectrum bands into one channel. Researching the SA technique in CR enabled TD-LTE system using TVWS is important.The task of this paper is from the national major projects for science and technology development of research and verification of Cognitive Radio in TD-LTE. The paper makes analyses and summary of relational issues, like Cognitive Radio, TVWS, TD-LTE, Femtocell and Spectrum Aggregation and its current researches and drawbacks. The innovative work of this paper is shown in the following two aspects:Traditional SA algorithms all assume that different secondary devices interfere with each other in all the spectrum fragments, so they allocate available spectrum fragments in an absolutely orthogonal way, which leads to the low spectrum utilization. In Femtocell, the interference relationships between Femto Base Stations (FBS) are nonuniform. As to it, the paper proposes a database-assisted spectrum aggregation to utilize the relationship. First, a database architecture control entity is introduced by which interference relationship between FBSs can be obtained in a cognitive manner. Second, graph coloring theory is used to model the system and optimization problem. Last, the labeling and coloring rule to solve the optimization problem and the steps of proposed algorithm are introduced. The simulation results show that comparing to traditional methods the proposed algorithm achieves better spectrum utilization by using the relationship between FBSs.Traditional SA algorithms consider only the aggregation and allocation of spectrum in only one sensing period aiming at maximizing spectrum utilization. It will lead to frequent spectrum switches between sensing periods, which will bring about a lot of controlling overheads. As for this problem, the paper proposes a spectrum aggregation algorithm for minimum spectrum switch. First, ON-OFF source and alternating renewal process are used to model the system and optimization problem. Second, the spectrum switching probability of spectrum aggregation region is calculated by alternating renewal process theory. Last, the steps of proposed algorithm are introduced. The simulation results show that comparing to traditional methods the proposed algorithm effectively reduces spectrum switches while guaranteeing the system access performance.