WIRELESS sensor networks (WSNs) have attracted tremendous attention for their mission-driven development and deployment. Improving Spectrum Efficiency via In-Network Computations in Cognitive Radio Sensor Networks For a large-scale WSN comprising lots of sensors, providing an efficient spectrum sharing with existing wireless networks is surely a trend. As facing the increasing spectrum demand of wireless services and devices , cognitive radio technology is widely employed to enhance spectrum utilization . Specifically, exploiting WSNs for smart grid applications , spectrum-aware technique is recognized as a promising solution to enable reliable Improving Spectrum Efficiency via In-Network Computations in Cognitive Radio Sensor Networks and low-cost remote monitoring for smart grids. To fully exploit this technology especially for large WSNs , more concurrent transmission opportunities within given spectrum are desired to realize spatial reuse of spectrum. In addition, maintaining reliable data transportation on top of numerous opportunistic links in cognitive (radio) multi-hop sensor networks becomes an essential requirement to bring the spectrum efficiency into reality. However, Improving Spectrum Efficiency via In-Network Computations in Cognitive Radio Sensor Networks as indicated by , there exists an significant end-to-end delay for greater network diameter in large cognitive machine networks and prevents practical applications. Thus, it becomes a great challenge to support an effective end-to-end quality-of-service (QoS) guarantee with regards of reliable communications in cognitive radio sensor networks (CSNs), while such likely technology is applicable for machine-to-machine communications, cyberphysical systems , and spectrum-sharing WSNs. Improving Spectrum Efficiency via In-Network Computations in Cognitive Radio Sensor Networks To achieve efficient spectrum management for cognitive radios, it is often done via forming the allocation optimization problems, such as spectrum or resource block allocation, user-based station assignment, and so on. Regarding multi-channel cognitive radio networks, time-spectrum blocks are allocated by constructing the subset of the good assignments and therefore obtain the suboptimal from given assignments Improving Spectrum Efficiency via In-Network Computations in Cognitive Radio Sensor Networks. A CSMA-based multi-channel MAC protocol is proposed by that optimizes the throughput performance for co-existing multiple systems. A distributed multi-channel MAC protocol is further proposed by for energy-efficient communication in multi-hop cognitive radio networks. Above efforts only focus on the efficient allocation of primary systems’ (PSs’) spectrum holes.