Revised DECT standard brings better voice in the IoT
It may not be as important as e = mc2, but for Digital European Cordless Telephony (DECT), the implications are potentially just as dramatic. As the technological universe migrates to both voice command and control and low-power Internet of Things (IoT) devices, the once European-centric cordless phone standard may be finding its way back into the global home after having been booted out by the rise of mobile phones.
This exploration of the future of DECT started with the question as to why Alexa’s Wi-Fi-based external call quality was so poor. This was quickly followed up by the questions: What if it were an emergency? And how could any Wi-Fi-based voice-enabled IoT device be trusted to operate effectively?
This raises questions for the whole field of home health monitoring and voice-based control, for both emergencies and casual, reliable, frustration-free use. The answer may lie with the addition of yet another wireless interface to the home particularly as voice-enabled devices and control become critical for the IoT.
The whole issue started rather innocuously: Using multiple Amazon Echo Dot 2s to make room-to-room intercom calls worked well over Wi-Fi. There were a few dropped syllables, but the conversations were fairly fluid. The problem arose when asking Alexa to make external calls: The connections were horrendous. Why?
It turns out that the Echo 2’s codecs are optimized to communicate with other Echo 2s over Wi-Fi, but not with the home router. In order to make reasonable quality external calls, it’s necessary to either sign up for a virtual VoIP PBX service skill or to purchase a $34 Amazon Echo Connect that connects any Alexa device to the home phone line for better-quality voice connections.
To be sure, Wi-Fi’s quality-of-service (QoS) enhancements have made it much better at giving voice priority over video or audio streaming where a few dropped packets are acceptable. However, each adaptation adds processing overhead, and not every Wi-Fi device is up to date with the latest standards. Also, Wi-Fi hardware was designed for high-speed data, and while it can be throttled back and use various sleep modes to lower power consumption, the hardware is relatively expensive.
However, DECT was specifically designed for high-quality voice at low power, not for high-speed data (though it can be used for data). In Europe, it operates in the 1,880- to 1,900-MHz band, while in the U.S., it is assigned to the slightly higher 1,920- to 1,930-MHz band to avoid interfering with public safety and government systems. All of the handsets connect back to a base station that connects to the phoneline.
Though still widely used in the enterprise, DECT lost ground in the home, squeezed between “good-enough” voice over Wi-Fi and mobile phones. Still, one of its salient features is its low latency of 10 ms, compared to 150 to 200 ms for Wi-Fi.
Given its low power and latency, among other features, it wasn’t surprising that in 2013, the ULE Alliance was formed with the goal of using a revised form of DECT to low-power, high-quality voice command and control and ensuring that IoT devices can communicate effectively and reliably. The move was prescient: Since then, voice control using Amazon’s Alexa and Google Home has taken off along with the IoT.
The target applications are home security, energy control, home automation, and medical and fitness wearables and emergency contact assurance. To address these markets, the Alliance’s changes to DECT mainly reside above the physical layer (PHY) and include: increasing the sleep time between beaconing the base station to save battery life; increasing security from 64-bit encryption to 128-bit AES; and the development of a star topology such that nodes connect directly to the base station, now called a concentrator. Communication to the outside and to other nodes all go through the concentrator.
>> Continue reading page 2 of this article on our sister site, EE Times: "Voice + IoT = DECT."