The Edge Computing paradigm and the Internet of Things

In 2018, there are more connected devices than humans on the planet. This impressive number is near 10,000 million and keeps growing at a high rate, with more than one-third of those devices being M2M (Machine to Machine), without human interaction. In this regard, more than 3,000 million WiFi devices were sold in 2017, and the services built around wireless communications generated the 4.4 % of the global GDP (Gross Domestic Product). Furthermore, in 2020, the number of public WiFi hotspots will fold 7x, reaching more than 430 million.

This trend will grow exponentially in the next years, thanks to the irruption of IoT (Internet of Things), where connected actuators and sensors are distributed all over the cities, transports, and industries. The business around this new paradigm has reached 700,000 million € at the end of 2017, and it is estimated that there will be more than 50,000 million devices connected to the internet in 2020. Half of them will correspond to IoT, like cameras, temperature sensors, intelligent illuminations, wearables, and more.

As the reader may have noted, IoT is driving networks toward dispersion. Edge computing is a new architecture in which substantial computing (resource optimization, data analysis) is placed at the network edge, in close proximity to devices. Thus, the data is processed adjacent to the point where it was generated. Tactile internet defines networks that combine low latency with high availability, reliability, and security. Together, a network implementing both paradigms is:

  • More secure: sensible data is not flowing through the network.
  • More robust: absence of bottlenecks and single-point-of-failure.
  • Faster: decisions are taken without delays.
  • More efficient: fewer data to the backbone, real-time resource optimization.

In this regard, most of the commercial WiFi solutions provide a cloud or centralized controller for the APs, avoiding the edge computing approach. This implies bottlenecks, delays in the decision-making, and critical nodes. Galgus has been working in this direction during the last 4 years, developing and patenting CHT (Cognitive HotspotTM Technology). CHT is an embedded software that, when installed, unleashes the true potential of APs (Access Points) and wireless routers.

CHT galvanizes your WiFi infrastructure, measuring the environment and sharing information between neighbor APs, promoting a fair use of radio and logic resources. In addition, the network administrator may build added-value applications on top, analyzing user’s behaviors, finding patterns, and exploiting contextual information without disturbing its users.

The WiFi standard was designed years ago around the concept of competition: devices fight to gain access to the resources, operating in an uncoordinated way. Our distributed intelligence lives at the edge of the network, so it is the best way to gather fresh information from the users, doing this way before than other cloud-based or centralized services. CHT does all the processing with a focus on efficiency and robustness, promoting a cooperative WiFi world (instead of competitive).

Ok, this edge computing approach sounds promising. But … which devices does Galgus sell? CHT is a pure software solution, so it works on a wide variety of WiFi devices. Just like Android is installed on smartphones. Stay tuned to this blog in order to know more about how Galgus product disaggregates software from hardware in WiFi networks.

 

Dr. Pablo Aguilera (Galgus Research)

 

 

Leave a Reply

Categories