Joint Scheduling and Power Allocation for 6G Terahertz Mesh Networks

Self-backhauling is a promising solution for ultra-dense networks. Since the Teraherz (THz) band can provide a very broad bandwidth, THz-based self-backhauling is an attractive technology for 6G ultra-dense networks to achieve “near-wired” transmissions and reduce cost. Based on THz, small base stations form a wireless mesh backhaul network, and a small base station can communicate with multiple neighbors concurrently. However, in a mesh network, time-slot must be carefully scheduled among directional links to fully exploit concurrent transmission opportunities while avoid cross-link interference and conflictions. Besides, since the THz band has frequency-selective and distance-selective properties, when allocating sub-bands to links with different transmission distance, the Long-User-Central-Window (LUCW) principle must be followed. Moreover, transmit power needs to be proper allocated to avoid interference. Therefore, the resource allocation problem in 6G THz backhaul networks is complicated. In this paper, a joint time-slot scheduling, sub-band scheduling, and power allocation (JTSP) scheme is proposed to solve the problem. This problem is formulated as a mixed integer nonlinear program (MINLP) problem, which is NP-Complete. To reduce the computational complexity, a greedy shrinking algorithm (GSA) is designed in JTSP to obtain a suboptimal solution. The effectiveness of JTSP is validated by simulations. Results illustrate that compared with traditional methods and algorithms, the proposed scheme can achieve 12.5%-60.7% throughput gain.