In the United States, ISM devices are regulated by the Federal Communications Commission (FCC) under Code of Federal Regulations 47 Part 15. Specifically, CFR 47§15.5 contains a general provision that devices may not cause interference and must accept interference from other sources. It also prohibits the operation of devices once the FCC notifies the operator that the device is causing interference. The commission has type-accepted (approved for use) 9,000 models of devices that can be sold and operate in the 902 – 928 MHz ISM band.
With these many devices operating without specific coordination or oversight, interference isn't just probable — it's palpable.
Spectrum networks, on the other hand, are private. Government regulators lease or sell use of an assigned bandwidth range that may be used only by a specific licensed user in a particular region. Interference isn't tolerated within that region, and is protected by government agency enforcement.INTERFERENCE
Mesh networks often have a high noise floor. A noise floor is like people talking during a movie; due to the number of voices being heard, understanding what's being said during the movie becomes more difficult. When industrial, institutional, and medical devices are all sharing spectrum, the noise floor is high. With more devices “talking” above, below, and even on the same operating frequency, utilities that deploy a mesh system architecture face challenges in the signal-to-noise ratio because the public spectrum receiver sensitivity is reduced, causing signal-to-noise ratios to fall. High signal-to-noise ratios are ideal because they offer superior throughput and reliability.
Mesh networks are vulnerable to high noise levels generated in shared frequencies. This reduces the ratio and provides inferior throughput and reliability, a critical factor in leak detection.
|AMERICAN NATIONAL STANDARDS INSTITUTE GUIDANCE|
C12: Smart Grid Meter Package
- Requirements and guidance on electricity metering, watt-hour meter sockets, device data tables, meter interfacing to data communication networks, and type 2 optical ports
- Performance criteria for thermal demand meters, mechanical demand registers, and phase-shifting devices used in metering • Test methods for transformer-rated meters and self-contained “A” base watt-hour meters are included in this package as well as a watt-hour safety standard.
X9.112-2009: Wireless Management and Security — Part 1: General Requirements
- Overview of radio frequency technologies
- General requirements applicable to all wireless implementations for the financial services industry.
A common tactic to combat low signal-to-noise ratios is to reduce the distance between a transmitter and receiver in which an access point/collector can operate. Within a dramatically smaller operating area, endpoints are physically closer and signal-to-noise ratios are increased. Even if noise is prevalent in an area, stronger signal-to-noise ratios can overcome interference but with a significant loss of range.
Another tactic is to exploit access to a wide or large band by either sending a direct sequence, high-speed transmission to disperse the signal across the band, much like a garden hose spray can be fanned out to a wide fast flow; or frequency hopping, in which the signal switches rapidly from one frequency to another according to a pseudorandom code.
Spectrum systems have a naturally low noise floor, maintaining excellent signal-to-noise ratios across larger distances and in the presence of signals on nearby bands. Like an open highway, signal traffic moves swiftly and travels further than when plagued by congestion.RANGE
Mesh networks are prohibited from generating more than 1 W of output, so their signals' range is limited. Even if they could transmit further, they would suffer poor signal-to-noise ratios across longer distances. For these reasons, mesh networks locate many points close together and move signals across a larger area through a series of short-range transmissions to intermediate nodes.
Licensed spectrum systems enable utilities to use higher power levels to optimize performance. Because of this flexibility, licensed spectrum networks are virtually interference-free and untroubled by crowded channels, as opposed to mesh networks whose power allotment largely relegates architecture to line-of-sight coverage only. Licensed spectrum signals routinely reach many times the distance of mesh signals.BANDWIDTH REQUIREMENTS
Mesh networks use a lot of bandwidth for each transmission because the data “hops” from node to node and requires a new slice of spectrum for each step. As such, the cumulative sum of bandwidth for sending a signal from its source to a final endpoint can really add up because you may have to buy more antennas and receivers.
Licensed spectrum works with a narrower band, but isn't as abundant. It's not free, either; but must be purchased or leased, sometimes in auctions where bidders must compete for licensed bandwidth.LATENCY
Unlicensed mesh networks involve a processing step with each node they reach, and this slows the signal's process to its destination. This lag or latency increases not only with distance, but when there is traffic from voice communications or a high volume of other data. Like interference, latency affects the ability to detect leaks.
Licensed spectrum systems allow signals to move through fewer or no midpoint nodes, so processing time is minimal and the signal moves swiftly to its destination. Low latency, or reduced delay time, may be increasingly important in the data-heavy 21st century.
—Ellis is the senior product manager, radio infrastructure and spectrum portfolio, for Sensus.