Communications with guaranteed bandwidth and low latency using frequency-referenced multiplexing
New cloud-based applications like VR and the connected vehicle fleet need secure connections and low and steady latency to the data centers at the edge. Current cloud communications for users rely on scheduled data frames via fiber networks with tree topology, which cannot guarantee connections with stable or low latency. They must also be more scalable to provide a broader range of users. We present a frequency-referenced multiplexing technique that ensures speed and low latency in time-sensitive applications. We employ optical and clock frequency synchronization enabled by frequency comb and signal processing methods to provide each user with a dedicated optical bandwidth, resulting in a scalable upstream user-cloud. In a demonstration, we can present a frequency-division-multiplexing system serving as many as 64 people, with a total capacity of 160 GHz with speeds that can reach 4.3 Gbps per user (240.0 Gbps total capacity when considering 200 GHz wavelength band) with the highest with a sensitivity of -35dBm.
The amount of VR (VR) devices, as well as Augmented Reality (AR) devices and intelligent autonomous vehicles, are predicted to increase rapidly in the coming decade (Fig. 1.), leading to a significant increase in the global volume of data 1. Contrary to the use of traditional information traffic (such as streaming video), the new cloud applications need secure connections (rather than a high-band rate) as well as high and reliable latency to data centers at the edge (Table 1) 3,4,5,6. The growth technology, therefore, has created distinct demands for the next generation of cloud-based user communication infrastructures linked to precision, availability, scalability, flexibility, and compatibility.
Fig. 1: Trends in the global growth of VR/AR devices and connected cars.
An estimated number of devices for human-machine interfaces specifically designed for VR and AR in the significant regions of the United States, Europe, China, and Japan. It is estimated that there will be a rapid growth of AR/VR apps, with an average growth (AAGR) of around 28% between 2021 and 2025 and an accelerating AAGR of about 14% between 2025 and 2030. The VR/AR devices are applicant time-sensitive applications like remote surgery and immersive education, teleconferences, online gaming, and experimental design 55. b, Estimations summarized as collected by PricewaterhouseCoopers and Strategy&56, showing an increase in connected cars in operation to 403 million by 2025, featuring an AAGR of 14% from 2021 to 2025, followed by an AAGR of 10%, reaching 645 million by 2030. These estimates represent the top geographical areas with connected vehicles, including those in the United States, Europe, China, and Japan.
Table 1 Future information rates and latency requirements for AR/VR devices and the connected vehicles
Tables of all sizes
In particular, the infrastructures must ensure that connections are high-quality and stable latency, in which the physical link between the cloud and the user should preferably be a specific channel so that uncertainties due to scheduling are eliminated, allowing precise time distribution through an infrastructure for telecom.
They must provide a precise and synchronized clock that allows the sub-nanosecond synchronization of time. Self-driving cars and VR require, for instance af, fordable and easily adaptable time synchronization infrastructures that can provide sub-metre-scale positioning for equipment seven. While sub-nanosecond time synchronization can be possible with GPS’, it’s expensive. It depends on the lines of sight towards the sky, which is often difficult or impossible to achieve, for instance, when in the underground or inside buildings 8 or 9 hindered by model atmospheric ash accumulated after volcano eruptions 9. Thus, sub-nanosecond time sync using the existing optical fiber access networks is highly sought-after.
The guaranteed connections will be scalable enough to support many user devices. The coming decade could have a quintuple rise in the number of devices used by users (Fig. 1.) This means that the future access networks should, at a minimum, offer the accessibility and accuracy required for a quintuple increase in users.