With the growth of mobile data traffic catalyzed by the world’s increasing dependency on cellular connections in every area of life, cellular networks are pushing capacity. However, spectrum is the scarcest resource in wireless technology and, with data usage skyrocketing, limited airwave access can be a bottleneck for growth and profitability. Just last year, mobile operators paid the U.S. government close to $100 billion for more 5G frequencies.
As spectrum is a finite resource and expensive investment, operators need to find ways to improve their use of existing network resources and cellular bands, especially as they are faced with growing Capex and a constant fight to gain new and keep existing subscribers. They can do this by investing in improving their networks by adding new hardware or cellular sites, or re-farming existing spectrum bands to enable new 5G services. This investment is required even if they have been lucky enough to gain new mid- or upper-band spectrum. Either way, yearly network investments alone will be between $5 and $20 billion, depending on the network operator and region.
Open Radio Access Network (O-RAN) developments promise to reduce the cost of building out a new 4G or 5G network, opening up the door for new vendors to compete for Capex spend. However, with new network vendors comes the added cost of integration. Operators are carefully evaluating the O-RAN approach to network programmability and the possibility of innovative services that may or may not be available from the incumbent vendors. For O-RAN to succeed, the ecosystem must be able to deliver new and differentiated slicing services that can appeal to existing subscribers and medium to large enterprises. It’s clear that the momentum for O-RAN is building.
However, while the O-RAN battle is being fought, operators still need to find a way to make more out of their existing spectrum assets, including the ability to deliver MU-MIMO and massive MIMO to FDD network bands, from 4T4R all the way up to 64TR.
A software-centric approach
While there will always be a need for traditional solutions to spectrum scarcity, such as investing in new cellular sites and base stations, there are alternatives that can have more immediate benefits. By adopting a software-centric approach to enhancing spectrum efficiencies, operators can maximize the efficiency of spectrum without having to wait for hardware upgrades to be made to existing handsets, radios or antennas.
The current software-based approach to spectrum sharing — Dynamic Spectrum Sharing (DSS) — enables 4G and 5G spectrum to be simultaneously deployed and co-exist within a single carrier, but it penalizes 5G performance to accommodate 4G. Operators therefore need to turn to software that is capable of supporting any waveform, any G, on any RAN (including O-RAN and its RIC xApps domain). This technology takes advantage of innovations in multi-user MIMO and beamforming to enable more efficient spectrum sharing between generations and spectrum multiplication within any G that works existing spectrum twice as hard to double the capacity at a single cell site.
The Delay-Doppler channel model enables MU-MIMO for 4G and 5G
When spectrum multiplier software is based on a Delay-Doppler channel representation, rather than the traditional time-frequency domain, the 5G performance issues seen with DSS can be mitigated. By superimposing time, frequency and — most importantly — space, the geometric Delay-Doppler model is slower changing and more predictable into the future than the conventional approach.
Due to the limitations of the time-frequency domain, up until now massive MU-MIMO hasn’t delivered fully on its promise. By assigning ample bandwidth to users connected to the same cellular site, Delay-Doppler software is fundamental to boosting cell capacity in future 5G networks and supporting the insatiable demand for more, and faster, connections. Fortunately, the use of the Delay-Doppler channel representation can change this by enabling true spatial multiplexing. By geometrically representing the channel across all of the user equipment (UE) attached to each base station, it enables simultaneous orthogonal beams to be formed to multiple UEs in the same spectrum — therefore, multiplying the capacity of the available spectrum.
Furthermore, the Delay-Doppler channel model provides an update to the existing time-frequency model as it is waveform and frequency independent. Consequently, this enables its benefits to be experienced in FDD and TDD spectrum and solve for both 4G and 5G congestion problems at once.
As mobile network operators battle to stay one step ahead of the demands of their subscribers, spectrum multiplication is a crucial weapon. Continuous spending on building out their existing 4G and 5G networks with new or re-farmed spectrum isn’t a viable long-term solution as the promise of 5G — which will help them recoup their spending — remains to be seen. This new era of spectrum multiplication based on a Delay-Doppler channel representation as opposed to the traditional time-frequency domain will allow operators to improve spectral utilization and performance for any waveform, now and in the future. This will not only allow them to continue providing their subscribers with the capacity and performance they expect, but will let operators begin to recover their costs in order to maintain profit margins while keeping up with strong network traffic growth.