Home

Motivation for Tree Splitting Medium Access Algorithms

With the advent of LTE and now, the imminent roll out of 5G, the number of devices sharing the same resources is growing at a rapid pace. These resources (Time, Frequencyand Code) need to be utilized wisely and hence arises the need for smart design schemes to solve the medium access problem. The first three generations of wireless cellular networks used FDMA, TDMA and CDMA respectively, which are fixed resource allocation protocols and hence are contention free by orthogonality of frequency, time and code. Even in these , at least the resource reservation channel is shared amongst the different nodes, and such a channel with many nodes has to deal with the collision of different packets. Resource reservation protocols are mainly limited by low channel utilization andlong delay in access when there is irregularly bursty traffic with many nodes. In such conditions random access protocols are more convenient. The main three types of random access protocols are ALOHA, CSMA and collision-resolution algorithms (CRA). The first two types suffer from stability and low throughput problems while CRAs can provide guaranteed resolution when the packets have collided and stable throughput. With many improvements in the early 2000s in digital satellite communication networks and cabled modems tree-based CRAs have gained interest once again. Examples where they are used are DAVIC/DVB, IEEE 802.14 and DOCSIS. In the family of CRAs the FirstCome First Serve (FCFS) 0.487 based on tree pruning and SICTA 0.693 based on the idea that collision of packets can be saved and later decoded, provide stability and high throughput. Collision resolution using successive interference cancellation techniques in signal shaping and processing have made SICTA-like algorithms a strong contender in the future of communication networks.