There is no single class of IoT device. They come in all shapes and sizes.
You designed your network for workstations and data communications. Next there was VoIP, which changed some of the network characteristics by demands to reduce delay and jitter. Then came video with higher bandwidth requirements and streaming. Now you have to embrace IoT, but there is no single form of IoT traffic to support.
IoT Traffic
The traffic created by IoT devices varies considerably. The IoT traffic will also depend whether or not an edge processor is used, how much intelligence it contains, and what applications are running on the edge processor. Smart edge processor intelligence can respond locally to sensors and process most of the locally data generated in the gateway. Dumb devices require more interaction over the network with the data center or cloud service.
Traffic characteristics can include:
- Continuous traffic flow--streaming
- Polled endpoints to collect data
- Periodic data with automatic delivery
- An IoT device may have actuators that require interactive communications
- Passive devices do not contain actuators, they only report sensors data
Delay tolerance is the amount of time a data transfer takes to reach a destination. When IP networks transfer data, delays of 100s of milliseconds was acceptable. When Voice over IP (VoIP) and Video over IP were introduced, 150 ms was considered the upper delay acceptable.
Delay tolerance for IoT divides depends on the application. If you are dealing with measuring the emptiness of a fuel container like a propane or oil tank, this application can easily tolerate seconds of delay. If the IoT endpoint is dealing with safety data or manufacturing systems, then 100 ms may be too long.
Data intensity is the amount of data created by a sensor. This can vary considerably. Very low bandwidth usage will occur when devices such as locks, doors, and light switches report their status. When an edge processor is used, then large volumes of data may be transferred that covers what the application creates. Video HD cameras and some industrial machinery also fall into the high bandwidth requirement category.
IoT Candidate Network Technologies
There is no shortage of technologies available for IoT networks. There are wireless and wired connections available. Within the wireless world, you have the existing 4G and the emerging 5G cell services. Most IoT devices don’t need high bandwidth of 5G networks. 5G is most attractive because of the number of IoT devices that can be supported.
A different technology is called low-power wide area networks. This is very good for covering areas which don’t have cell service or in rural areas or where the cost may be excessive for the cell services.
Wi-Fi is possible for small environments. Bluetooth is possible but most likely will not be used because of the distance and bandwidth limitations.
A new addition to the 5G networks is called Citizens Band Radio Service (CBRS). This is competitive to Wi-Fi and can offer longer distances and higher bandwidth. CBRS is an emerging technology. There are not that many products available yet but expect them in the future.
The only reason for mentioning satellite services is that there will be places like Rocky Mountains states, Alaska, and much of Canada that will not have any available wireless service other than the satellite.
The wired world has an interesting capability which is operating power over Ethernet over twisted pair category 1, 2, and 3 cable. This requires either cable adapters or a specifically designed local-area network switch. This reduces the wire requirements. You use existing telephone cables. This solution can operate over 1000 to 2000 feet of cable.
Classic category 5 or higher Ethernet is also available but remember that the cable distance limitation is 100 meters or 330 feet.
In the case of video cameras and surveillance equipment, coaxial cable is a candidate. It can offer both high-speed and long distance over several thousand feet.
Fiber-optic cable is possible but likely to be too expensive for the large number of IoT endpoints that need to be connected.
Edited by
Ken Briodagh
