Echo cancellation plays an important role in current Internet protocol（IP） based voice interactive systems. Voice state detection is an essential part in echo cancellation. It mainly comprises two parts： double talk detection（DTD） and voice activity detection（VAD）. DTD is used to detect doubletalk and prevent filter divergence in the presence of near-end speech, and VAD is used to determine the near-end voice activity and output silence indicator when near-end is silent. However, DTD straightforwardly proceeded may mistakenly declare double talk under double silent condition, coefficients update under the far-end silence condition may lead to filter divergence, and current VAD algorithms may misjudge the residual echo from the near end to be far-end voice. Therefore, a voice detection algorithm combining DTD and far-end VAD is proposed. DTD is implemented when VAD declares far-end speech, filtering and coefficients update will be halted when VAD declares far-end silence, and the far-end VAD adopted is multi-feature VAD based on short-time energy and correlation. The new algorithm can improve the accuracy of DTD, prevent filter divergence, and exclude the circumstance that far-end signal only contains residual echo from near end. Actual test results show that the voice state decision of the new algorithm is accurate, and the performance of echo cancellation is improved.
This paper deals with network selection problem for users in heterogeneous network environment. The main context is to improve the TOPSIS( Technique for Order Preference by Similarity to Ideal Solution) network scheme by combining the network properties and the users' requirement accurately and decrease ping-pong effect. The method of entropy and FAHP( Fuzzy Analytic Hierarchy Process) are used to calculate weight value and the sojourn time calculation is used to avoid ping-pang effect. The simulation results show that the improved scheme enhances the more accuracy of network selection than the existing methods and reduces the number of ping-pang effect.
User pairing strategy for virtual multi-input multi-output （VMIMO） has been widely studied to improve system throughput, but most studies are based on perfect channel state information （CSI） and uniform power allocation. However, perfect CSI is very difficult or even impossible to obtain in practical system. Moreover power allocation has significant impact on algorithm performance. Therefore, in this paper, a low-complex joint user pairing and power allocation algorithm based on aggressive discrete stochastic optimization and Lagrangian dual solution is proposed for uplink VMIMO with imperfect CSI. Simulation results show that the proposed algorithm can achieve desirable throughput performance, and restrict inter-user interference （IUI） efficiently.
This paper focuses on the energy efficient relay selection problem in a cooperative multi-relay network,aims to find the most energy efficient relay node for the source node while ensuring its minimum data rate requirement.The interaction between the source node and the relay nodes is modeled as a Vickrey auction game,when the source node broadcasts a cooperation request,the relay nodes compete for the cooperation,and the one with the minimum bid will be chosen which denotes the cost of the source node during the cooperation process,but it only needs to provide the minimum bid provided by the other relay nodes,which can encourage all the relay nodes to give the true bid.Besides,the minimum rate requirement of the source node will be ensured and the relay node taking part in the cooperation will gain some reward and the reward can be maximized by reinforcement learning(RL).
Orthogonal frequency division multiplexing （OFDM） which has been adopted in the long-term evolution （LTE） system can improve the system capacity obviously. However, it also brings about severe inter-cell interference （ICI） for cell-edge users （CEUs）. To tackle this problem, multi-user selection and power control （MuS-PC） is proposed as an efficient scheme in uplink coordinated multi-point multi-user multi-input multi-output （CoMP-MU-MIMO） transmission/reception. This paper jointly considers user＇s signal to interference plus noise ratio （S1NR） and proportional fairness （PF） to maximize the total channel capacity in multi-user selection by formulating a penalty function. To simplify the penalty function＇s computation, particle swarm optimization （PSO） algorithm is introduced. In addition, power control is adopted to maximize overall energy efficiency. Simulation results demonstrate that the MuS-PC scheme can not only obtain the optimal total channel capacity while guarantee each user＇s quality of service （QoS） and PF, but also largely reduce computational complexity and improve energy efficiency. As a result, the poor communication quality of CEUs can be enhanced.
Network selection and resource allocation( NS-RA) are the processes of determining network and radio resource which provide the service to user. Optimizing these processes is an important step towards maximizing the utilization of current and future networks. In this paper,we proposed a preference value-based network selection and resource allocation,in which the NS scheme was performed by the joint radio resource management( JRRM) entity and the RA scheme was performed by the network. In the NS step,the JRRM entity selected the preferable network for users according to the preference value of each network,which took the load balance,the received signal strength( RSS) and the relative position between the user and the network into account. In the second step,the network allocated the optimal sub-carrier to user for the downlink transmission each round according to the preference value of each user and the maximum reachable data rate calculated by users' perceived channel information,maximizing the spectrum efficiency as well as guaranteeing the fairness. The simulation results showed that the proposed NS-RA scheme achieves better performance in terms of load distribution,spectrum efficiency and user fairness,compared to the conventional strategies.
In downlink coordinated multi-point（CoMP） system, full cooperation is always not applicable in real world because of its high request in the backhaul. To deal with this problem, clustering decision is made to process transmission. In this paper clustering methods based on the metric signal-to-leakage-plus-noise（SLNR） is proposed. In addition, user scheduling schemes based on SLNR is also put up to make the scheduling set as large as possible. Simulation results show that the proposed clustering methods not only reduce the data sharing among the cooperating base stations（BSs）, but also improve the system throughput compared with the traditional clustering methods based on channel strength.
Financial and environment considerations present new trends in wireless network known as green communication. As one of the most promising network architectures, the device-to-device （D2D） communication should take seriously account to the energy-efficiency. Most of the existing work in the area of D2D communication only focus on the direct communication, however, the direct link D2D communication has to be limited in practice because of long distance, poor propagation medium and cellular interference, etc. A new energy-efficient multi-hop routing algorithm was investigated for multi-hop D2D system by jointly optimizing channel reusing and power allocation. Firstly, the energy-efficient multi-hop routing problem was formulated as a combinatorial optimization problem. Secondly, to obtain a desirable solution with reasonable computation cost, a heuristic multi-hop routing algorithm was presented to solve the formulated problem and to achieve a satisfactory energy-efficiency performance. Simulation shows the effectiveness of the proposed routing algorithm.
Due to the constraint of single carrier frequency division multiple access （SC-FDMA） adopted in long term evolution （LTE） uplink, subcarriers allocated to single user equipment （UE） must be contiguous. This contiguous allocation constraint limits resource allocation flexibility and makes the resource scheduling problem more complex. Most of the existing work cannot well meet UE＇s quality of service （QoS） requirement, because they just try to improve system performance mainly based on channel condition or buffer size. This paper proposes a novel resource scheduling scheme considering channel condition, buffer size and packet delay when allocating frequency resource. Firstly, optimization function is formulated, which aims to minimize sum of weight for bits still left in UE buffer after each scheduling slot. QoS is the main concern factor here. Then, to get packet delay information, this paper proposes a delay estimation algorithm. Relay node （RN） is introduced to improve overall channel condition. Specific RN selection strategy is also depicted in the scheme. Most important of all, a creative negotiation mechanism is included in the subcarrier allocation process. It can improve the overall system throughput performance in guarantee of user＇s QoS requirement. Simulation results demonstrate that the scheme can greatly enhance system performance like delay, throughput and jitter.
The dynamic resource allocation problem in high-speed railway downlink orthogonal frequency-division multiplexing（OFDM） systems with multiple-input multiple-output（MIMO） antennas is investigated.Sub-carriers,antennas,time slots,and power are jointly considered.The problem of multi-dimensional resource allocation is formulated as a mixed-integer nonlinear programming problem.The effect of the moving speed on Doppler shift is analyzed to calculate the inter-carrier interference power.The optimization objective is to maximize the system throughput under the constraint of a total transmitted power that is no greater than a certain threshold.In order to reduce the computational complexity,a suboptimal solution to the optimization problem is obtained by a two-step method.First,sub-carriers,antennas,and time slots are assigned to users under the assumption of equal power allocation.Next,the power allocation problem is solved according to the result of the first-step resource allocation.Simulation results show that the proposed multi-dimensional resource allocation strategy has a significant performance improvement in terms of system throughput compared with the existing one.
