A page replacement algorithm based on a fuzzy approach to improve cache memory performance

The memory management in the operating system includes a part called the page replacement algorithms. Replacement algorithms in environments that require high-performance computing are considered as an important issue. For example, these algorithms are very important in cache management in microprocessors, web caching, replication strategies in distributed information systems and so on. Due to the important role of replacement algorithms in overcoming the problem of performance caused by the difference in processor speeds and memory, many algorithms were proposed. Most of them are the developed schemes of the least frequently used (LFU) and least recently used (LRU). Although most proposed designs can solve the LRU and LFU defects, they are implemented in a difficult way. The most important advantage of LRU and LFU is their simple implementation. This research proposes a page replacement algorithm that is simple to implement. The algorithm, which uses three parameters to cluster cache pages, is called the fuzzy page replacement algorithm. Whenever a miss occurs, it selects a page of the cluster with the lowest priority which has the smallest Euclidean distance with its center and then exits the cache. The most significant advantage of the algorithm is using the FCM (fuzzy c-means) algorithm to cluster pages, resulting in better replacement and hence higher memory performance.