Machine Learning to Improve Success Rates for Handovers from Sub-6 GHz LTE to Millimeter Wave Bands
Transmission over millimeter wave (mmWave) frequency bands is being adopted in fifth generation (5G) wireless communications. Even though the sub-6 GHz frequency bands continue to dominate deployments due to their better ability to penetrate and provide in-building coverage, the handover between mmWave and sub-6 GHz frequency bands is nonetheless inevitable to support higher data rates. The cost of a handover is a reduction in data rate, which 5G promises to increase.
While it is always possible for handset manufacturers to introduce new receiver circuitry for mmWave bands in 5G, the price of such a handset will be more expensive and the battery will deplete faster when having to measure two bands simultaneously.
WNCG Professor Brian L. Evans and PhD student Mr. Faris Mismar have introduced a new category of partially blind algorithms for inter-radio access technology handover between LTE-Advanced and 5G. The handover is partially blind because it makes a decision to handover to mmWave bands based on the current sub-6 GHz LTE-Advanced channel measurements. The decision comes from the probability of handover success from an extreme gradient boosting classifier that has been trained using simultaneous LTE-Advanced and mmWave channel measurements. If the classifier finds that the handover to mmWave is likely to succeed, it will allow the handover measurements to start by allowing the LTE-Advanced base station to configure an occasion for the handset to measure the mmWave channel. This occasion is known as the measurement gap. If not, the base station instructs the handset to remain in the LTE-Advanced service area. This way, the measurement gap is not opened, and the handset continues to receive scheduled data from the better performing technology.
F. B. Mismar and B. L. Evans, “Partially Blind Handovers for mmWave New Radio Aided by Sub-6 GHz LTE Signaling”, Proc. IEEE International Conference on Communications Workshop on Evolutional Tech. & Ecosystems for 5G Phase II, May 20-24, 2018, Kansas City, MO, USA.