The priority queue model divides CR (cognitive radio) system users into primary user and secondary user. The primary user has preemptive priority over secondary user, while the same priority level users are FIFO served. This paper models the CR system based on scheduling technology. The secondary users are further divided into different priority levels. A hybrid priority dynamic policy, which indicates primary user's preemptive priority and secondary user's nonpreemptive priority, is developed to reduce spectrum switch overhead during spectrum leasing process. Theoretical analyzing shows that hybrid priority reduces the spectrum switch overhead and the hybrid priority improvement is a non-negative mono-increasing function of priority level. Based on the proposed policy, CR scheduling model is built and CR scheduling rule is provided and proved. Experimental comparison between CR scheduling and priority queue shows that CR scheduling gets a smaller weighted staying time sum than priority queue and the improvement increases as system becomes more congested.

Spectrum leasing has made more efficient utilization for the limited spectrum resource. While increasing the spectrum utilization, spectrum leasing also brings in some problem. Spectrum switch in the leasing process not only introduces in additional system overhead, but also decreases the secondary user's Qos. In this paper, we model the wireless system as a M/G/1 queue with preemptive priority, propose a new channel parameter to analyze the switch performance. Algorithms are suggested to reduce system overhead and to improve secondary user's Qos based on the new parameter.

In the wireless network of finite spectrum resource, some communication users probably do not allocate the spectrum so that their communication should be delayed. Thus it is necessary to arrange the communication time of the systems when we allocate the communication frequency. We construct a frequency-time spectrum model and propose a joint pre-allocation in a two-dimension space. Through the distributed games learning algorithm, each system could obtain its communication schedule, which makes the bandwidth utilization achieve a quite high level and arranges the global communication time as short as possible. The experiments show that our results have obvious superiorities in both bandwidth utilization and temporal performance relative to other allocation algorithms.

In cooperative multi-agent reinforcement learning (MARL), communication can address the challenges of partial observability and environmental non-stationarity by conveying environmental features and decision intents, respectively. However, existing methods either focus on only one type of information—failing to tackle both challenges simultaneously—or conflate these signals, causing agents to confuse environmental context with decision intents. This paper introduces Disentangling Environment and Decision messages for Multi-Agent Communication (DEDMAC), a framework that explicitly separates these two information types into two distinct message streams and processes them independently. Specifically, environment messages are integrated into long-term memory to resolve partial observability, while decision messages provide instantaneous intent signals to mitigate non-stationarity and facilitate coordination. To prevent semantic confusion between the two message streams, we employ mutual information constraints to ensure semantic disentanglement. Furthermore, we design a mechanism that leverages global information to correct intent biases in decision messages resulting from limited local perspectives during generation. Evaluations across complex multi-agent benchmarks demonstrate that DEDMAC significantly outperforms state-of-the-art communication-based methods. These findings indicate that the explicit separation and specialized processing of environment and decision semantics are critical for achieving optimal performance in dynamic, collaborative multi-agent systems.