Practical Considerations for Wireless Sensor Networks and Mesh Networking
In Crossing the Chasm, Geoffrey A. Moore introduced the Technology Adoption Lifecycle. This curve highlights a chasm that exists between visionaries who are early adopters of technology and pragmatists who wait for proven technologies and products. For technologies to “cross the chasm” one key component are products with defined product specifications and system deployment guidelines.
Wireless Sensor Networks (WSNs) are an emerging technology that has reached a critical point in the Technology Adoption Lifecycle where the success of innovators and early adopters combined with products from proven vendors will allow end users to begin to address challenging deployments. This process is underway and National Instruments is providing products with defined system performance and familiar graphical software development tools.
Different Wireless Technologies
First to understand the capabilities of WSN systems we need to understand the technology. There are several different options for wireless systems including licensed technologies such as cellular and unlicensed technologies in the 900 MHz, 2.4 GHz, and 5 GHz range. In the 2.4 GHz unlicensed band there are two standards that are commonly used, IEEE 802.11 b/g and IEEE 802.15.4. The first step is to understand the trade-offs in bandwidth, range and power.
Bandwidth, Range, and Power
The IEEE 802.11 and IEEE 802.15.4 protocols have distinct differences. Wi-Fi based on IEEE-802.11 has the advantage in bandwidth with a maximum bit rate of 54 Mbit/s for IEEE 802.11g; IEEE 802.15.4 which is used in several protocols including ZigBee has the advantage in distance and power requirements. Wi-Fi offers significantly higher data rates, which require additional encoding; extra data requires additional radio traffic resulting in increased power consumption by the radio. This bandwidth and power trade-off is obvious in systems such as laptops or smart phones with integrated Wi-Fi that typically operate for a matter of days, compared to a wireless sensor network based on IEEE 802.15.4 technology that might operate for years on standard AA batteries.
Defined Bandwidth and Measurement Performance
One of the steps in the transition from early innovators to pragmatic applications is the availability of products with defined specifications and system performance benchmarks.
For high-speed wireless measurements, National Instruments offers Wi-Fi DAQ. One application is high-speed machine condition monitoring which requires high accuracy vibration measurements. For a Wi-Fi DAQ module with 4 channels and a sample rate of 51.2 kS/s per channel the required throughput is 6.6 Mbit/s. For comparison, a common broadband data rate in the US is 6.5 Mbit/s.
As an example of system performance for IEEE 802.15.4 products, National Instruments offers NI WSN products. The throughput required for a remote monitoring application with 64 channels from 8 wireless nodes connected to a single gateway is 5.2 kbit/s. In this case each node offers 4 analog and 4 digital I/O channels and a 1 second sample interval.
A NI WSN system can expand to a few thousand channels by adding additional WSN gateways on new IEEE 802.15.4 channels. The maximum number of channels in a system will depend on the network traffic in the 2.4 GHz range and the network topology.
Range and Power Requirements
Although the wireless streaming bandwidth advantage with Wi-Fi is clear, technologies based on IEEE 802.15.4 offer advantages with range and power requirements. For example, the National Instruments Wi-Fi DAQ products offer a range up to 100 m line of sight and operate with 9-30 VDC inputs. In comparison NI WSN products can support up to 300 m line of sight and can operate at 1 sample per minute for up to 3 years. Wireless ranges are dependent on the application environment and a site survey can help ensure correct topology of wireless devices.
Another area is network topologies. A Wi-Fi system is typically configured in a star topology with a center Access Point and Clients 30 to 100 m from the access point depending on the wireless environment. While standard Wi-Fi installations support repeaters or routers to extend distance with a tree topology, they do not support mesh networking. Mesh networking is the ability for a node or device to route packets through multiple paths back to the gateway, like in Figure 2.
The NI WSN platform supports powered mesh networking which offers two benefits:
- Ability to extend the maximum distance from end node to gateway
- Back-up paths for environments where signal loss is a concern
For products that implement mesh networking, today they either use powered routers or a network timing scheme. Products that use powered routers enable routers to quickly forward messages from end nodes when the node has data ready to send. This leads to a wireless network that is more responsive to several end nodes and will better support asynchronous measurements. An alternative approach is to add timing to the network so the entire network will wake at pre-set time intervals. This has the advantage of not requiring power for the routers; however in certain uses an end node could consume more power if it is required to wake more frequently than is necessary for the measurement requirements. The network responsiveness depends on the time interval and a short time interval can reduce the power efficiency.
Graphical System Design Software for Wireless and Wired Measurements
For pragmatists to deploy WSN systems one important component are products with defined specifications and system benchmarks. With benchmarks engineers can design WSN systems to address applications including environmental, asset, energy and water monitoring and with Wi-Fi measurement systems engineers can address challenges in structural health and machine condition monitoring.
In addition to hardware products, engineers also need familiar graphical software design tools like LabVIEW for wireless system configuration and deployment in addition to acquisition, storage, and analysis tools for wireless measurement data. With these types of open software tools customers can combine wired and wireless measurement systems, connect to third party wireless measurements and WSN networks from a wide range of vendors and even deploy graphical programs to NI WSN nodes to customize their behavior.