The term 5G — meaning the fifth generation of mobile network or wireless systems — is appearing more and more these days. There's talk of initial deployments as early as 2020. What is less talked about is the fact that the 5G standards are still being considered and defined, and many aspects are still in flux.
So what does 5G entail? Well, the Next Generation Mobile Networks (NGMN) Alliance defines the following requirements that a 5G standard should fulfill:
- Data rates of tens of megabits per second for tens of thousands of users.
- Data rates of 100 megabits per second for metropolitan areas.
- 1 Gb per second simultaneously to many workers on the same office floor.
- Several hundreds of thousands of simultaneous connections for wireless sensors.
- Spectral efficiency significantly enhanced compared to 4G.
- Improved coverage.
- Enhanced signaling efficiency.
- Significantly reduced latency.
In a nutshell, 5G will provide a unified protocol bringing together a massive number of users with requirements as diverse as MTC and eMBB (enhanced Mobile Broadband); it will support new use cases that LTE cannot handle such as VR, V2X, eHealth, and mission-critical emergency services; it will be compatible with the previous generation LTE and LTE-A Pro network infrastructure (Core and RAN); and it will seamlessly aggregate very diverse frequency bands from 400MHz to 80GHz with bandwidths from 20 to 800 MHz using LTE, LTE-A Pro, 5G NR, and Wi-Fi 11ax/ad with a unified protocol. (This all seems so simple if you wave your hands around and say it quickly.)
Consider V2X for example. This encompasses a variety of scenarios, including V2V (vehicle-to-vehicle) and V2I (vehicle-to-infrastructure). In the case of V2V, autonomous automobiles will be communicating with each other in real-time saying things like “I'm almost at the intersection so please slow down a little to let me through.” Obviously, this type of communication requires high fidelity and low latency.
The end result is that 5G will be a wireless standard that's so advanced compared to previous generations that it will demand extreme processing to ensure its success.
According to ABI Research, 5G mobile data traffic is expected to represent 40% of total data traffic and have 500 million subscribers by the year 2025. What's really interesting is the fact that 3G and 4G usage is expected to continue to grow linearly, while 5G is expected to increase in an exponential fashion.
Mobile data traffic by technology (Source: CEVA/ABI Research)
Not surprisingly, 5G presents a host of computing challenges. Technologies designed for wireless standards like 5G will need to be capable of delivering a peak data rate of up to 20 Gbps under ultra-low-latency of 1 millisecond. This will be achieved utilizing innovative and extremely complex processing techniques such as Massive-MIMO and advanced 3D dynamic beamforming. DSP processors deployed for today’s LTE-Advanced Pro and multi-gigabit wireless standards are simply not capable of efficiently delivering the speed, latency, and overall DSP performance required to address the massive technology leap to 5G.
All of which explains why CEVA has just announced its CEVA-XC12 DSP core, which offers the raw performance and power efficiency that is fundamental to the success of multi-gigabit-class modems. Thanks to its flexible architecture with multiple optional features, the CEVA-XC12 can be custom-configured and scaled to address a wide range of applications. This includes smartphones and other terminals, advanced and centralized access points, small cells, macro cells, and cloud RAN (C-RAN).
The CEVA-XC12 also supports the full gamut of 5G use cases and deployment scenarios, from 80 GHz mmWave down to 450 MHz spectrum bands. In addition to 5G, the CEVA-XC12 is well-suited for the design of LTE-Advanced Pro Evolution, enhanced Mobile Broadband (eMBB), Licensed Assisted Access (LAA), MulteFire carrier aggregation and LWA (LTE/Wi-Fi Aggregation), cellular V2X, Wi-Fi 802.11ax, WiGig 802.11ad, Fixed Wireless Access (FWA) and Virtual Reality (VR) systems.
The CEVA-XC12 DSP architecture is underpinned by six key technologies as follows:
- A new micro-architecture to meet very high frequency requirements and ultra-low power consumption — capable of operating at 1.8 GHz in 10 nm and 50% less power than its predecessor, the CEVA-XC4500.
- Massive computation capabilities to maintain a high bit-rate — equipped with quad-vector processor engines approaching 1 tera operations per second (TOPs) performance.
- New and unique high-precision arithmetic — achieves optimal resolution with up to 256×256 dimension matrix processing.
- New specialized instructions to boost all baseband processing components — innovative support for advanced 256 and 1024 QAM demodulation.
- New core streaming interfaces — allowing ultra-low latency transfers between cores or accelerators.
- New control plane for massive-user management and for multi-RAT (Radio Access Technology) systems — incorporates a Scalar Processing Unit with a CoreMark/MHz score of 4.4 designed to handle huge number of users required for LTE MTC and 5G IoT.
The CEVA-XC12 also features a state-of-the-art cache-architecture and support for hardware coherency for seamless multi-core implementations. The CEVA-XC12 DSP core is available for licensing now (click here for more information).