Scale & Integrate Disparate Building Management Systems
Wireless Sensor Networks
Guidelines for the use of wireless sensor networks in building applications.
Adaptive Wireless Solutions Ltd
Meeting government targets for energy reduction requires significant
effort in understanding the use of energy in buildings, measuring
consumption in new and retrofitted properties and communicating data to
Wireless communication technology opens up a wealth of opportunities for monitoring and controlling conditions within a building by easing the retrofit installation of sensors and control outputs. Wireless technology enables increased numbers of sensors, actuators, and controllers in a building by drastically reducing the cost, effort and disruption of installation. The elimination of signal wire also provides greater flexibility within spaces with adaptable configurations.
Despite the apparent ease with which people can deploy wireless sensors and actuators in a building, engineers and operators still have concerns and questions regarding the reliability of wireless technology.
To address these concerns potential users of wireless technology require clear guidelines on how to design systems for a range of building types, common factors that are present and a design methodology which should ensure the maximum chance of avoiding problems once the system is deployed in the field.
There are a number of standards in the wireless sensor network domain. The common denominator is the IEEE 802.15.4 specification which defines the lowest layers in the communications stack. The higher layers in the stack are generally defined by individual vendors who focus on specific performance issues or bodies responsible for applications in specific vertical markets e.g. ISA, Wireless Hart in the process industries.
Because of the generic nature of wireless sensor technology it is likely that a number of standards are likely to exist in the market place driven by the demands of each market. Potential users need to be careful to avoid blindly pursuing a particular "standard" simply because the promoters of the standard have achieved some success in raising the profile generally in the market place. It is safest to consider standards as simply one of a number of equally important factors necessary for successful wireless applications.
There are a number of traditional wireless solutions which fall in to the category of point-to-point or point-to-multipoint (or ‘star’) networks. The reliability of these networks is set by the quality of the RF link between the central access point and each endpoint. In many application settings it can be difficult to find a location for an access point that provides dependable communications with each endpoint. Moving an access point to improve communications with one endpoint will often degrade communications with other endpoints.
Wireless mesh networks communicate with each other through routers and a gateway to form a mesh network. Most mesh networks use end nodes which are the primary sensing devices which relay their signals through router nodes (which may also act as sensors) and the router forwards the signal to a gateway device. All the mesh devices relay signals for each other, so if a given sensor is out of range of the gateway, other members of the network carry data.
Each vendor may implement their network architecture in different ways whilst using the basic design described and supported sensor types are also vendor specific.
The gateway device generally communicates with a higher level network for which each vendor provides a standard interfaces - e.g. Modbus, BACNET.
Factors influencing range include transmit and receive power, antenna design, operating frequency, building topology and obstructions.
Transmit power is regulated as part of the standards process whilst receive sensitivity continues to improve as designers and radio chipset manufacturers employ increasingly novel techniques to maximise reception whilst minimising power consumption.
Sensors may use external antennas to improve reception sensitivity. The decision is often based on other factors in the application including topology and other sources of potential interference.
Devices operating at 2.4 GHz appear to offer the bandwidth required to achieve the optimum combination of range and reliability for internal applications. Additionally the ISM band at 2.4 GHz provides the benefit of a worldwide standard whereas other frequencies have regional restrictions on their use.
Line of sight (LOS) distance is a metric that all radio product manufacturers cite as a performance metric. It is easy to quantify that metric when there are no obstacles
between the two measured points and it is extremely consistent.
The distance metric between wireless nodes in open air with line-of-sight conditions depends on radio frequency, transmit power and receive sensitivity. Battery-powered sensing devices are also often designed to operate at lower radio transmit power to conserve batteries. For 2.4 GHz systems based on the IEEE 802.15.4 standard, LOS range between battery-powered sensing devices and the mesh devices is typically 100-200 metres but can be up to 600 metres in some cases. Between line-powered mesh devices LOS range is more usually 400-600 metres. If any two nodes have an obstacle between them the operational distance will be reduced by a factor. The factor is variable depending upon the type of materials the obstacle is constructed from and the thickness and density of those materials.
Interference is a potential concern since it may lead to dropped or corrupted data.
Protocols employing a frequency hopping design are generally better at reducing the possibility of interference. Often the application design requires limited transmit times which also helps minimise the potential for interference. Because the possibility of lost data can never be totally eliminated there is a requirement for the application software to apply error checking rules as well as relying on the error checking and packet acknowledgement and retry rules built into the wireless network protocol.
Wireless network design
Network architecture is determined by the application location type. One approach is based on classifying building types as - commercial, light industrial or heavy industrial. Each type has common features in terms of building structure, material and space usage which affect the overall approach to network design.
Various best practices are emerging as wider experience with wireless network deployment is gained. A typical set of "rules" recommended by one vendor include:
Similarly best practices are emerging for the location of network devices - gateways, routers and sensor nodes.
Once overall network design and device locations are determined the network must be configured. Generally this must be carried out in the manner prescribed by the vendor as each is a little different.
In point of fact many mesh systems now do not require the installer to perform any manual wireless configuration steps other than either defining in advance the node ID’s to be part of the network or selecting them from a list of automatically ‘discovered’ devices. Once all devices are present the mesh network is configured automatically as part of the standard operation of the mesh network protocol.
The lifetime of battery powered sensors is an issue of common concern. To maximise battery life, wireless vendors carefully program their battery-powered nodes to minimize energy consumption by going to “sleep-mode” when not transmitting or receiving data.
The frequency of data acquisition and transmission, the power levels at which the radio transmitters are designed and the network topology (range) required, will affect the life of the battery-powered devices in the system.
In most applications line power will be available for powering repeater devices. In effect the network will consist of a mix of battery and line powered devices with totally wireless data transmission.
Experience shows that solely battery powered repeaters generally provide poor battery life but they can be used when a means of recharging is available (e.g. solar panel or intermittently-powered lamp post). Evidence from current installations has shown that a minimum battery life of 12 months can be expected from battery-powered sensing devices. This is very much a variable figure as battery technology, sensor design and network design are continually advancing.
Energy harvesting devices provide the benefit of eliminating the need for batteries. It is important to be aware of the amount of energy that must be collected to permit the data acquisition and radio transmissions that are needed. Also where energy harvesting based devices are employed there will generally be the need to provide additional battery or line powered devices to complete the network design.
Wireless sensor networks are a dynamic technology impacting multiple application areas. Success in applying the technology is based on applying sound engineering principles and using a structured process in order to address the issues which are common to all designs as well as those that relate to the specific application.
Design guidance is evolving as experience is gained in the field and as further successful installations are implemented.
About the Author
David Laurence is responsible for marketing and sales development at Adaptive Wireless Solutions Ltd. (www.adaptive-wireless.co.uk)
He has over 30 years of experience in the building controls, industrial automation and IT industries where his roles have included sales, engineering and management.
David’s main focus at present is concerned with how to enable the adoption of wireless sensor networking by understanding the issues and challenges faced by users in specific vertical markets. He holds a BEng in electronic engineering from Sheffield University.
[Click Banner To Learn More]
[Home Page] [The Automator] [About] [Subscribe ] [Contact Us]