Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 105 === With the rapid development of wireless broadband networks and mobile communication technologies, a wide range of mobile devices have been popularized. In recent years, because of the increasing density of mobile devices, diversified vehicles and application services have made users become accustomed to using mobile devices to handle life, communication, entertainment and other aspects of life.
However, the data signal and control signals must be transmitted through the core network, which will cause the congestion of the core network.
Therefore, device to device (D2D) can transmit signals to each other over a direct link using the cellular resources instead of through a base station base station has been paid more and more attention gradually.
In general, device-to-device increases spectrum utilization, overall throughput, and power consumption. However, due to the limited spectrum available, bandwidth utilization and distribution are more important.
When D2D communication under the underlay mode resource allocation, because of the relationship between reusing the resource block, the allocation of resource blocks to cellular network users and D2D users, there will be signal interference problems. Therefore, this thesis focuses on how multiple D2Ds schedule when they compete with each other at the same time to ensure that the resources are properly and fairly allocated.
In order to solve the above problems, this thesis proposes to share the resource blocks by combining the Hungarian algorithm with the proportional fair algorithm. The Hungarian algorithm will choose the maximum weight matching, which can increase the system throughput. Proportional fair algorithm will pulled high the D2D weight when the D2D data rate drops. Therefore, the D2D with low transmission rate has a greater chance to grab the resource block. By combining the advantages of the two algorithms, our proposed algorithm compared to others can balance the D2D fairness without sacrificing overall throughput.
|