Nokia and Vodafone collaborate on active antennas to boost 5G radio capacity and reduce costs Go back »
2019-02-24 | All chapters
Real-world network measurement data verifies ability of Machine Learning to boost 5G massive MIMO capacity and RFIC based massive MIMO active antenna applicability to urban macro coverage.
Espoo, Finland - At Mobile World Congress 2019, Nokia and Vodafone will demonstrate two massive Multiple Input Multiple Output (mMIMO) innovations to improve 5G capacity and performance. The companies will jointly showcase how 5G services can be delivered using low power, compact form factor and more easily deployed RFIC based mMIMO antennas, which reduce Total Cost of Ownership (TCO) compared to conventional mMIMO antennas. They will also demonstrate an industry-first Machine Learning powered scheduler for 5G massive MIMO, which can improve radio resource management and boost 5G radio performance.
Massive MIMO active antennas are essential for achieving the very high capacity potential of 5G radio. However, deploying the initial Massive MIMO products might be challenging on typical busy base station sites due to their size, weight and power consumption. Furthermore, optimizing the operation and performance of multiple antenna elements is extremely complex.
Vodafone and Nokia have jointly conducted extensive simulation studies concluding that the RFIC1based Nokia mMIMO active antenna is a more suitable solution to those challenges than the other radio solutions like for example using passive antennas and remote radio heads. The RFIC based Nokia mMIMO active antenna is an industry first solution to implement mass-manufactured terminal technology for more flexible, lower impact mMIMO antenna solutions. Vodafone will be the first mobile operator in Europe to field-test the Nokia active antennas incorporating RFIC.
Rapid mobile data traffic growth and 5G's ability to prioritize different performance parameters, such as capacity, coverage, reliability and latency, make optimization and radio resource management (RRM) a constantly changing challenge. Machine Learning (ML) algorithms reduce the complexity as well as the amount of processing capacity and other resources needed.
The ML-powered mMIMO scheduler for RRM and optimization can substantially improve 5G Radio Access Network (RAN) performance by optimally allocating radio resources (e. g. large numbers of antennas/beams/layers) to users and simplify the operation of 5G mMIMO. Nokia has developed the ML-powered scheduler and supporting architecture in collaboration with Vodafone, which provided relevant anonymized real-world data from its network. The scheduler solves the highly complex task of mMIMO scheduling in real-time to provide superior 5G user experiences.
The fruitful collaboration between Nokia and Vodafone started a year ago, initially focusing on enhanced RRM using ML for Channel Quality Indicator (CQI) optimization. This has now reached a proof of concept using Nokia commercial hardware and leveraging analysis of anonymized data provided from Vodafone network.
Please click the link below to read more.
Source: Nokia
