Taking It to the Edge: 5G Edge Computing Antenna Requirements
At hundreds of acres apiece and packed with AI workloads, hyperscale data centers dominate headlines. That makes it easy to overlook something that plays a fundamental role in supporting hyperscale AI workloads: edge computing.
As its name implies, edge computing puts processing resources close to where data is collected, acted on, or both. One reason is that proximity can enable faster decisions and actions. For example, edge computing can analyze data from nearby roadside sensors to identify sudden changes such as an accident and then push alerts out to vehicles and digital signage in the vicinity.
Another reason is that not all data needs to be shipped to a hyperscale data center dozens or hundreds of miles away. Edge computing can analyze streams coming in from IoT sensors to determine which data is worth the bandwidth to send to the hyperscale facility for AI analysis, whether that’s in real time or forensically. One example is surveillance video that is worth relaying only when there are incidents for AI to act on and learn from.
Cutting the Cord
Copper and fiber can link edge computing to IoT endpoints and to hyperscale data centers, but cellular is increasingly becoming the preferred option. One reason is because it’s easier and faster to deploy than trenching or stringing cable.
Another reason is that cellular keeps getting faster, both in terms of bandwidth and latency. That’s closing the gap with copper and fiber and making cellular a viable option for bandwidth-intensive and latency-sensitive edge computing applications that wouldn’t be possible with 4G.
5G supports peak speeds of up to 20 Gbps, which is 20 times faster than LTE-A, and latencies as low as 1 millisecond versus 30-70 msec for 4G. 6G is expected to support up to 100 Gbps, which is more than enough for even the most demanding applications, such as live 8K video. (For more information, see “4G vs. LTE vs. 5G: How Mobile Technology is Evolving” and “What is 6G?”)
In the case of 5G, another reason is maturity and the module savings that come with it. 5G made its commercial debut in 2019, so it’s had over six years to ride down the cost curve. IoT applications are notoriously price sensitive, especially with mass-scale deployments such as tens or hundreds of thousands of sensors and controllers around a smart city.
Many edge computing applications are low bandwidth or use battery-powered endpoints. 5G RedCap caters to both. Named for being a “reduced capability” version of 5G, RedCap is designed for devices that need 100 Mbps or less to send and receive data. RedCap also maximizes battery life with mechanisms such as discontinuous reception (DRX), where the device puts itself to sleep to save power. That’s ideal for IoT devices that need to remain in service for several years. (For more information, see “With 5G RedCap, Less is More for IoT.”)
Top Design Considerations
Although 5G provides a solid foundation for edge computing, the success of an edge computing gateway ultimately depends on how well engineers build on that foundation. Band selection and MIMO orders are top antenna considerations, as is PCB placement in the case of embedded antennas.
The application, technology, and module choices collectively determine the bands, MIMO order, and antenna choices. For example, if RedCap is the ideal choice for a particular application, only two cellular antennas are required for receiving (2×2) rather than the four that are typical for a full 5G system.
The target countries/regions determine the module and in turn which bands the antenna must support. If the solution will be sold globally, a wideband antenna is a good choice because it’s flexible enough to cover a broad range of public and private operators’ spectrum licenses. This helps enable single-SKU products and future proofing.
In the case of embedded antennas, an example is the Taoglas PCS.62.A, which covers 600 MHz to 6 GHz in an ultra-compact footprint of just 38 x 10.3mm. If external mounting is required, one good choice is the Taoglas Comet MA322, a puck-style antenna covering 600 MHz to 6 GHz, with two 5G/4G high-performance elements for MIMO. The MA322 is adhesive/magnetic mount, which is ideal for installations where drilling isn’t practical or possible. The Comet also is available in screw-mount versions.
The Taoglas PCS.62.A has a PCB keep-out area that’s only slightly larger than its footprint. That’s noteworthy because with any embedded antenna, the keep-out area directly affects performance. So when comparing embedded antennas, check each model’s integration guide for the recommended keep-out distance from battery cans, screws, cables, and copper traces, all of which can undermine radiated signals. (For more tips, see “Four Top Reasons Your GNSS and Cellular Antenna Performance is Suffering.”)
Although RedCap includes features that maximize battery life, the antenna can take that power efficiency to the next level. Unlike GNSS, cellular antennas can’t use active electronics such as Low-Noise Amplifiers (LNAs). That means thermal and power management are critical for minimizing power consumption and maximizing battery life.
This highlights the role of antenna efficiency. Poor antenna efficiency forces the transceiver to operate at higher power levels to maintain connectivity, thus increasing overall power consumption. But a well-designed antenna reduces the amount of transmit power necessary to maintain a reliable connection and in turn extends battery life beyond what RedCap itself enables.
Get in touch for orders or any queries: sales@rfdesign.co.za / +27 21 555 8400
Courtesy of Taoglas
