Cell-free network architectures emerge as a prospective solution for Terahertz (THz) communication systems by deploying dense access points to simultaneously serve user equipments, which shorten communication distances and enhance link reliability. Meanwhile, abundant bandwidth and narrow beams of THz links improve the channel capacity and reduce interference in cell-free networks. However, the resource allocation problem in THz cell-free networks is yet an open issue, especially regarding long-term optimization in varying wireless environments. In light of this, a dynamic resource allocation problem established on dynamic array-of-subarrays hybrid beamforming is proposed to maximize the long-term sum-rate in THz cell-free ultra-massive multiple-input multiple-output (UM-MIMO) networks. Given the non-convex nature of this optimization objective and the requirement of dynamic programming, a multi-task based, multi-agent deep reinforcement learning algorithm (MTADRL) for resource allocation is developed. Simulation results validate that the THz cell-free network enhances the data rates compared to small cell and cellular counterparts. Furthermore, MTADRL demonstrates robust convergence, achieving a 19% and 46% improvement in long-term sum-rate over the benchmark DRL and equal allocation schemes with a total transmit power of 20dBm and 64 sub-arrays, respectively.

Supporting ultra-high data rates and flexible reconfigurability, Terahertz (THz) mesh networks are attractive for next-generation wireless backhaul systems that empower the integrated access and backhaul (IAB). In THz mesh backhaul networks, the efficient cross-layer routing and long-term resource allocation is yet an open problem due to dynamic traffic demands as well as possible link failures caused by the high directivity and high non-line-of-sight (NLoS) path loss of THz spectrum. In addition, unpredictable data traffic and the mixed integer programming property with the NP-hard nature further challenge the effective routing and long-term resource allocation design. In this paper, a deep reinforcement learning (DRL) based cross-layer design in THz mesh backhaul networks (DEFLECT) is proposed, by considering dynamic traffic demands and possible sudden link failures. In DEFLECT, a heuristic routing metric is first devised to facilitate resource efficiency (RE) enhancement regarding energy and sub-array usages. Furthermore, a DRL based resource allocation algorithm is developed to realize long-term RE maximization and fast recovery from broken links. Specifically in the DRL method, the exploited multi-task structure cooperatively benefits joint power and sub-array allocation. Additionally, the leveraged hierarchical architecture realizes tailored resource allocation for each base station and learned knowledge transfer for fast recovery. Simulation results show that DEFLECT routing consumes less resource, compared to the minimal hop-count metric. Moreover, unlike conventional DRL methods causing packet loss and second-level latency, DEFLECT DRL realizes the long-term RE maximization with no packet loss and millisecond-level latency, and recovers resource-efficient backhaul from broken links within 1s.

Terahertz (THz) band communications are envisioned as a key technology for 6G and Beyond. As a fundamental wireless infrastructure, THz communication can boost abundant promising applications. In 2014, our team published two comprehensive roadmaps for the development and progress of THz communication networks, which helped the research community to start research on this subject afterwards. The topic of THz communications became very important and appealing to the research community due to 6G wireless systems design and development in recent years. Many papers are getting published covering different aspects of wireless systems using the THz band. With this paper, our aim is looking back to the last decade and revisiting the old problems and pointing out what has been achieved in the research community so far. Furthermore, in this paper, open challenges and new research directions still to be investigated for the THz band communication systems are presented, by covering diverse topics ranging from devices, channel behavior, communication and networking, to physical testbeds and demonstration systems. The key aspects presented in this paper will enable THz communications as a pillar of 6G and Beyond wireless systems in the next decade.

We consider a many-to-one sensor network where a large number of sensors are deployed to monitor a physical environment. We explore sensor activity management to maximize the network lifetime while meeting the quality of service (QoS) requirement. Specifically, in each round the sink estimates the number of active sensors and the control information is fed back to the sensors for activity control. We start with a basic case where the total number of sensors N is known, and the estimator of the number of active sensors fit is accurate. We devise a sensor activity control scheme under which the number of active sensors would converge to the minimum that can meet the QoS requirement. Next, we generalize the study to the following two more complicated cases. (1) The case with known N and inaccurate nt. For this case, we propose a stochastic approximation algorithm to minimize the average number of active sensors while meeting the QoS requirement. (2) The case with unknown N and accurate nt. For this case, we cast the problem as the adaptive control of a Markov chain with unknown parameters and propose a composite optimization-oriented approach for the corresponding sensor activity control. We show that using this composite optimization-oriented approach the number of active sensors would converge to the minimum that can meet the QoS requirement

In this paper, we consider a many-to-one sensor network where a large number of sensors are deployed to monitor a physical environment. We explore sensor activity management to maximize the network lifetime, while meeting the quality-of-service (QoS) requirement. Specifically, in each round the sink estimates the number of active sensors and the control information is fed back to the sensors for activity control. We start with a basic case where the total number of sensors N is known, and the estimator of the number of active sensors n/spl circ//sub t/ is accurate. We devise a sensor activity control scheme under which the number of active sensors would converge to the minimum that can meet the QoS requirement. Next, we generalize the study to the following two more complicated cases: (1) The case with known N and inaccurate n/spl circ//sub t/: For this case, we propose a stochastic approximation algorithm to minimize the average number of active sensors while meeting the QoS requirement. (2) The case with unknown N and accurate n/spl circ//sub t/: For this case, we cast the problem as the adaptive control of a Markov chain with unknown parameters and propose a composite optimization-oriented approach for the corresponding sensor activity control. We show that using this composite optimization-oriented approach the number of active sensors would converge to the minimum that can meet the QoS requirement.