An Integrated QoS Metric for Handoff Decisions in 5G and LTE Networks
Date
Authors
Sandhya Rani Siva Raju, Srinivas
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Growing demand for wireless data and the spectrum congestion in the existing 4G LTE networks have led the researchers to focus on ways to utilize the large bandwidths available in the above 6 GHz spectrum. The technology that would be implemented in this spectrum has been termed as 5G is expected to roll out by the year 2020. While many models have been proposed and researched on the implementation of 5G, relatively less work has been done on one of the crucial aspects of cellular communications: handoffs. Handoff is the process of transferring connection of a mobile user from one node to another, when the previous node cannot provide proper service. Current handoffs are performed, based on the signal strength. In this process, as the signal from the current node drops, the user then has to connect the node to a higher signal strength. While this method could work in micrometer wave spectrum, the propagation characteristics of a signal are different in the millimeter wave spectrum. Frequent drops in the signal level, even due to the user movement would mean that number of handoffs performed would be very high and would also lead to the ping-pong effect. To avoid this, the use of an integrated QoS metric for the handoff decision has been proposed. In this method the handoff is performed not based on the signal strength level but by considering past values of the each base station has been made. For the weighting criteria the normalized variance of SINR values of each base station, and the probability of missed detection has been considered. This ensures that the mobile user would connect to a more stable base station thus avoiding frequent unnecessary handoffs. A simulated approach has been taken to compare the performance of the weighted handoff with the traditional signal strength approach based handoff. The simulations were done in a 5G radio propagation framework using Matlab and in the LTE model of the Network Simulator 3 (NS3) software. The results show that the number of handoffs performed were lesser in the weighted handoff as compared to the traditional handoff, which was observed in both the 5G propagation simulation in Matlab and in the LTE model of NS3. In addition to this, analysis on the effects the weighted handoff would have on different network QoS parameters-the mean delay, the packet loss ratio and sum of jitter, has been done on NS3 LTE. Results show that the new algorithm doesn't just abide by the LTE QoS, but also shows an improvement.
Description
Keywords
Handoff decisions, Simple Additive Weighting (SAW), MmWave propagation, NS3