Research On Massive Machine Type Communications Technology For 5G

With the rapid development of Internet of Things(IoT),machine type communications(MTC)is being expected.The conventional cellular systems designed for human type communications can no longer satisfy the communicational demands of MTC.Massive MTC(mMTC)is one of the typical usage scenarios of the fifth generation of mobile communications(5G),and its features include large number of users,small packets,low user data rates,and sporadic user activity.Based on these distinctive features,uplink grant-free non-orthogonal multiple access(NOMA)scheme is highly expected to be applied in mMTC for the reduction of signaling overhead and transmission latency.Due to the sporadic nature of mMTC,the multi-user detection problem is similar to the reconstruction of sparse signals in compressive sensing theories.The signal reconstruction method of CS have a broad application prospect in the next generation of mobile communications.This paper starts with the single-slot sparsity model in the mMTC uplink grant-free NOMA scenario,focusing on the multi-slot sparsity model,including structured sparsity model,mixed sparsity model and dynamic sparsity model.In some typical usage scenarios of mMTC,structured sparsity exists in user activity.However,current compressive sensing based multi-user detection schemes assume the channel state information(CSI)remain constant in continuous time slots,which is unrealistic in 5G mMTC.Hence,a structured-compressive-sensing-based multi-user detection under the scenario of timevarying CSI is proposed to exploit the inherent structured sparsity of user activity,namely structured sparsity orthogonal matching pursuit(SSOMP)algorithm.Simulation results show the feasibility of the proposed algorithm in dynamic CSI framework.On the other hand,considering the temporal correlation of active user support sets in continuous time slots in dynamic sparsity model,a prior-information aided adaptive subspace pursuit(PIA-ASP)algorithm is proposed based on subspace pursuit(SP)algorithm.Then,to mitigate the incorrect estimation effect of the prior support quality information,a robust PIA-ASP algorithm is further proposed,which adaptively exploits the prior support based on the corresponding support quality information in a conservative way.Moreover,for the two proposed algorithms in dynamic sparsity model,the upper bound of the signal reconstruction error and the computational complexity is derived.Simulation results demonstrate that the two proposed algorithms are capable of achieving much better performance than that of the existing CS-based multi-user detection algorithms with a similar computational complexity.