چكيده به لاتين
5G wireless networks utilize massive MIMO technology to serve numerous high-rate Internet-of-Things (IoT) devices. The main challenge of massive MIMO technology is estimating and tracking the channels of devices. Since there are many devices in the IoT applications, it is impossible to assign a unique pilot to each device permanently due to the length of the coherence time. Thus, the use of massive MIMO systems has become a significant issue. The previous approaches cannot provide service for users with weaker signals. The previous channel tracking methods are limited to the fact that the BS should be aware of the pilot access pattern of users in each access slot. Additionally, the maximum number of users who can transmit the pilot to the BS is limited in all these methods since increasing the number of devices increases the probability of error and access latency.
This dissertation derives some novel methods to estimate and track the channels of high-rate IoT devices and also diminish the intra-cellular pilot contamination using data association concepts. Pilot access patterns of users in coordinated access should be notified to the BS. This leads to a reduction in system flexibility. Therefore, this dissertation offers a new viewpoint called uncoordinated access, in which all users can independently choose their pilot access slots at will without any coordination with the BS. Since it is ambiguous as to which users exist in the
collision per access slot, the channel estimation and tracking challenge become more complex
than coordinated access. In this dissertation, BS associates the same pilot to a group of devices.
Our proposed approaches not only resolve pilot collisions but also exploit their information. In addition, they explicitly account for channel aging resulting from users’ movement respective to
the BS.
The investigations conducted in coordinated access are as follows: 1) assuming a fixed channel state model, the optimum algorithm and two suboptimal trackers are derived. To reduce complexity, the optimum algorithm is redesigned based on the vector and matrices with smaller dimensions. Also, its steady-state feature is analyzed, 2) when users’ channels are under uncertain evolution, some analytical models are assumed for them. Then, channel tracking algorithms and their stability are investigated, and 3) a practical scenario is introduced in which the estimation of time-varying channels is not based on the assumption of analytical models. Also, the correlation between the additive noise to the received signal and the channel state noise is considered. The optimum algorithm and several suboptimal trackers are developed for channel tracking in this scenario.
The investigations conducted in uncoordinated access are as follows: 1) assuming a fixed channel state model, the optimum algorithm and six suboptimal trackers are proposed, 2) the near-optimal algorithm and several new trackers are designed with significantly lower complexity and better performance, 3) A new flexibility is added to uncoordinated access in
which all devices are allowed to freely select their individual activity possibility as well as pilot
access patterns. Then, channel estimation is analyzed, and 4) the users’ channels are tracked in uncoordinated access with the assumption of a mathematical model for estimating their uncertain evolution.
