In the architectural debates that rage around how 5G will look, opinions differ on whether a new air interface will be required (or several), or how far the next generation will just be an evolution, or amalgamation, of current technologies. But few see it in such dramatic terms as Marcus Weldon, president of Bell Labs, where so many of the telecoms industry’s R&D breakthroughs have been made. The mobile network is so close to the limits imposed by the laws of physics that 5G must perform no less than “magical things”.
He told a recent open day at the Labs’ Irish facility near Dublin: “The physics of access are now so constrained that we need to do magical things. 5G things.” The executive, who is also CTO of Bell Labs’ owner Alcatel-Lucent, said the R&D unit had calculated the amount of room left to improve access before hitting the barriers imposed by physics. The figures are alarming – improvements of just 1.4 times are left using spectral efficiency and x5 by harnessing millimeter wave spectrum, which many tout as a centrepiece of 5G.
The potential is greater when it comes to deploying ever-smaller cells, an architecture which could still deliver 18-fold improvements. Weldon is a believer in densification as a way to squeeze more capacity out of spectrum and recently introduced a Bell Labs report, which concludes that a dense HetNet could deliver 50% energy savings, 10% greater uplink throughput and 50% better downlink. However, those improvements are only achievable if inter-cell interference is convincingly addressed using techniques like eICIC in LTE-Advanced. Weldon said: “Findings from the project underscore the critical importance of small cells and prove that that by deploying a mix of base station radios operators will see better energy efficiency and increased network performance – thereby improving their customers’ experience and increasing efficiency.”
However, Weldon told the Dublin event, dubbed FutureX Days, that small cells would have their limits too – raising the specter of a dedicated cell for each user, with antennas added to the humans themselves. Not that such a vision is inconceivable. Artemis, the start-up behind the pCell (personal cell), is pushing an architecture which would, in effect, create a temporary cell of capacity around a user when required. The system harnesses, rather than mitigates, the interference between signals and uses that interference to synthesize cells of just one centimetre in diameter, centered on the antenna of a standard LTE device, which deliver the same data rates as a full base station by sharing spectrum channels. The firm describes its architecture as ‘software defined radio Cloud-RAN’.
This sounds futuristic, and has been much-hyped, but it is getting the chance to show off its paces in two trials – one a test network created by base station vendor pureWave in 700MHz Band Class 14 public safety spectrum in Silicon Valley; the other in Dish spectrum in California.
There are plenty of other efforts to push the boundaries further in terms of spectrum usage and mobile capacity – whether emanating from universities, vendor and operator labs or start-ups. Several focus on full-duplex transmission (sending and receiving on the same frequencies at the same time, long deemed impossible). One of these efforts comes from Columbia University in New York, whose researchers have demonstrated a CMOS chip they call CoSMIC (Columbia high-Speed and Mm-wave IC) which can perform full-dupex and so double the speed of a wireless communication.
The university’s professor of electrical engineering, Harish Krishnaswamy, told EETimes: “Full-duplex at the same frequency has been thought to be impossible because if you transmit and receive at the same time at the same frequency, the receiver will get drowned out by the interference from the transmitter, which can be 1bn times more powerful than the signal that one is trying to receive. But we have devised an echo- or self-interference cancellation scheme that can do away with the transmitter’s interference with one part per billion accuracy, so that after cancellation, the received signal can be easily detected.” The team is targeting wireless, particularly cellular and WiFi, but says the technique could also be used to double the capacity of optical channels.
Also promising commercially usable full-duplex systems is Kumu, a start-up spun out of rival university Stanford in Silicon Valley. It came to the world’s attention last year when it demonstrated incoming and outgoing signals using the same frequency without interference, effectively doing away with the FDD/TDD difference. Then it found a way to incorporate its technology relatively easily into existing base stations, potentially paving the way for near term market adoption, rather than having to wait for the next wave of wireless systems.
Kumu’s Steven Hong told the ThinkSmallCell newsletter that “there is no DD anymore – not FDD or TDD”. The firm says it has completed proof of concept trials successfully with four tier one operators and will run field trials, using equipment from major vendors, this year, with commercial deployment in 2016. It is no secret that two of those operators will be Telefonica and SK Telecom – at Mobile World Congress, both were demonstrating the Kumu system on their stands.
The company addresses the interference issue by connecting a smart echo-canceller in place of the usual RF rx/tx combiner/splitter in a standard base station. The Kumu radio generates an additional signal that cancels out the interference – a different approach to established interference cancellation techniques, and one which could be applied to cellular networks or to WiFi.
According to Kumu VP of product development Joel Brand, the system is constantly scanning the radio environment, gauging the state of the airwaves at any given time, as well as the way the transmission is changing the signal environment at the receiver. Its algorithms then compensate for those changes as the signals arrive from the opposite direction, in effect applying echo cancellation to radio waves.
As well as improving speeds, Kumu says spare macrocell capacity created via this approach could be harnessed for backhauling nearby small cells, dividing the full 100Mbps of capacity in the macro sector between several metrocells, which would not need their own backhaul, and could also act as relays. In this scenario, the small cell would need to implement the special receiver but no special handsets or macrocells would be required. Another potential, and topical, application is to combine WiFi and LTE-U (LTE using 5GHz carriers for supplemental capacity) in one box while avoiding the usual interference issues.
Initially, however, Kumu will also be targeting its invention at backhaul, an area where it believes it can get the quickest wins because of the need for new capacity and the less standardized hardware environment.
The same approach is seen at another full-duplex specialist, MIMOtech, which is also focusing on backhaul first. The company says it can double backhaul throughput in microwave bands using its Air Division Duplexing (ADD) technique, which combines full-duplex with MIMO. At MWC, the company launched its Janus AirDuplex range of microwave backhaul radios based on this technology, claiming data rates of up to 1Gbps in a single 28MHz channel, and 2Gbps in 56MHz.
CEO Buks Fouche said the system is more cost-effective than using millimeter wave spectrum above 42GHz for backhaul – an approach which many are testing for small cells because of the plentiful capacity available. He said: “The wireless industry is facing a crisis in its ability to keep up with the demand for backhaul capacity, and Janus AirDuplex provides a way to achieve higher throughput without resorting to using the millimeter wave frequency bands above 42GHz, which offer broader bandwidths but suffer atmospheric attenuation and also often attract higher licensing fees.”
ADD uses spatial multiplexing to double the capacity and spectral efficiency of conventional microwave platforms. With an antenna separation of about 100 millimeters, it can be considered as, in effect, a single-aperture antenna from the point of view of licensing, site rental cost and implementation, says its inventor, reducing capex and opex costs.
Backhaul is also the key focus of another network innovation, though this one is around the modulation not the duplexing. Magnacom – which is backed by former Qualcomm president Steve Altman, the architect of that company’s licensing model – emerged from stealth mode last year with its WAM (Wave Modulation) technology, a scheme which founders Yossi Cohen and Amir Eliaz believe could be an alternative to QAM, boosting speeds and reducing power consumption.
They say WAM is a drop-in replacement for QAM circuitry, using the same I/Q interfaces, and does not require any changes to analog or RF components. Nevertheless, it promises an improvement of up to 10dB in system gain when compared to QAM4096, which in turn, the inventor says, results in up to 50% greater power and spectral efficiency, higher speeds and a fourfold lengthening of range. Plus, a simpler design should lead to lower costs.
So there is no shortage of innovation and radical thinking going into what will become 5G. But these changes will take a long time to develop from exciting demoes in Barcelona to affordable, mass market equipment. This is where Weldon gets the last word, issuing one of his repeated warnings against 5G hype. He told the Dublin event: “The problem is we are five years away from it and we have already exhausted the hype. That’s generally not so good because it means we have to talk about something else soon. Then you are devaluing 5G.”