|
چکیده
|
Millimeter wave (mmWave) communication is considered as a promising technology to support high speed data services in wireless networks. Motivated by the huge available bandwidth, mmWave networks can support bandwidth-intensive delay-sensitive applica tions such as multimedia streaming. However, the conventional resource optimization schemes in the traditional wireless networks will not be efficient for multimedia trans mission in mmWave networks due to the severe path loss and high blockage probability. In this paper, the problem of dynamic joint flow control, channel assignment, and power allocation is studied for multimedia transmission in multi-connectivity based mmWave networks. To fully exploit the mmWave communication potentials, a concurrent transmis sion scheme is proposed to improve the spatial spectrum reuse that results from directional communication. Multiple simultaneous access point (AP) connections are adopted to ad dress the potential blockage in mmWave networks. The goal of the considered stochastic optimization problem is to maximize the aggregate utility while minimizing the power consumption and preserving queue stability subject to the long-term minimum data rates constraint. By leveraging Lyaponuv technique and given the channel allocation decision, the problem is decomposed into the flow control and power allocation problems. Then, using dual decomposition technique, it is shown that the power allocation problem can be implemented in a distributed way via an exchange of demand and price among the net work and users, whereby each user solves an unconstrained convex optimization problem. Simulation results demonstrate that a tradeoff between the utility performance and queuing delay can be achieved.
|