| Summary: | 碩士 === 國立雲林科技大學 === 資訊工程系 === 105 === Toward 5G, 3GPP specifies the cooperative communictions in Long Term Evolution-Advanced (LTE-A) to forms a cooperative set of eNBs or cell sections for each UE. The user-centric cooperative transmissions easily achieve data sending via multiple transmission paths while reducing interference, channel quality, data rate, etc. Clearly, for a UE near cell edge still anchoring at the original serving eNB before performing handover, the cooperative communication certainly reduces transmission delay and increases reliability. Although such a user-centric cooperative communication can achieve these advantages, it is an open issue in 3GPP specifications. Moreover, several critical issues exhibit in cooperative communications, including 1) how to dynamically form the set for a moving UE, 2) high computing complexity for RAN (Radio Access Network) to compute the cooperative set for each UE, 3) when to initiate the cooperative communication, 4) significantly increasing RB allocation cost, etc. Thus, this paper proposes the UE heading Prediction Optimal Cooperative set selection approach (POC) that consists of three phases: 1) the phase of QoS-based Priority Function (QPF), 2) the phase of Adaptive UE Heading Prediction of moving path (HP), 3) the phase of Adaptive Optimal Cooperative mode determination and RB allocation (AOC). The main contributions include: 1) adaptively determining the cooperative set for each UEs with the UE heading prediction mechanism, 2) accurately determining the optimal cooperative communication mode, 3) maximizing the system capacity and reward. Numerical results show that the proposed approach outperforms the compared approaches in system capacity, access delay, average dropping probability, reward and net-profit. Furthermore, the computation complexity of the proposed approach is analyzed, i.e.,T(n)150 = O(0.07233*n^3), which is less than that of the compared approaches without any predictive heading of a moving UE, i.e.,T(n)150<T(n)360 = O(n^3).
|
|---|
