Communication Systems

Impacts

Developing a Custom Point-to-point Link Protocol

Point to Point Protocol

There are many international standards that define RF communication protocols for a variety of applications. They range from the highly complex cellular standards that define the operation of our 2G, 3G and 4G networks to the more modest standards, such as 802.15.4. The simplicity and flexibility of 802.15.4, in particular, underpins systems such as ZigBee, WirelessHART, and many others.

And yet despite this choice, there is still a call for custom protocols to meet specialised requirements. Custom protocols are ideal for applications that are cellular but have endpoints travelling at higher speeds than supported by cellular standards. Another application is when the endpoints are all stationary but very low power consumption is needed.

A recent requirement from one of our clients was a point-to-point system in which the radio links are subject to severe shadow fading and multipath Rayleigh fading. Each link was to act as a substitute for an Ethernet wire, providing over 100 Mbps of throughput under good conditions and, more importantly, high availability low-rate communications under poor conditions.

The first step, as with any project, is to determine the requirements. Both the direct requirements specified by the client and the indirect requirements that arise from a detailed analysis and understanding of the client’s application. In the case of a protocol, the RF environment is one of the areas that need to be most carefully studied, as are the regulations that will apply to its use in different parts of the world.

An early decision was to treat the Ethernet replacement requirement literally. This meant that the user data transported by each link would consist of Ethernet frames, and the control data exchanged between endpoints would only be used for managing the link itself. Higher layer control and configuration for the client’s network would be transported as IP data and could, therefore, be secured with established security solutions. The result was that the protocol to be developed would only need to define a Physical and Medium Access Control layer.

After careful consideration, Orthogonal Frequency-Division Multiplexing (OFDM) was chosen because of its flexibility. Under good conditions, the RF channel can be packed with a large number of sub-channels, each modulated with a high order modulation, e.g. 256 QAM. Under poor conditions, the transmit power can be concentrated in just a few sub-channels and the modulation order can be dropped to 4 QAM.

Extensive simulations were carried out to confirm the suitability of the choice given OFDM’s two principal disadvantages: its susceptibility to receiver frequency errors and the need for highly linear signal chains, particularly in the transmitter. The simulations included power amplifier models and synthesizer phase noise models, as well as models of the fading environment.

The Medium Access Control layer imposed a Time Division Multiplexing structure on the transmitted signal that was designed to simplify the implementation of the client’s receivers while ensuring the system’s ability to support a range of propagation conditions. High-end throughput was sacrificed to this end, but a peak throughput of over 130 Mbps was nevertheless achieved.

Convolutional codes were used for forward error correction, and adaptive link control adjusted the number of active sub-channels, the modulation, and the code rate so that a wide range of throughputs could be supported, ensuring that the high availability requirement could be met.

It is rare that a communication system needs a custom low-level RF protocol to be developed and it is an expensive process. However, it can open up applications and services that would otherwise not be possible.

Benefits of a Proprietary Solution

Contact Us – Email: hello@plextek.com or call: +44 (0) 1799 533 200

Developing a Custom Point-to-point Link Protocol

Point to Point Protocol

There are many international standards that define RF communication protocols for a variety of applications. They range from the highly complex cellular standards that define the operation of our 2G, 3G and 4G networks to the more modest standards, such as 802.15.4. The simplicity and flexibility of 802.15.4, in particular, underpins systems such as ZigBee, WirelessHART, and many others.

And yet despite this choice, there is still a call for custom protocols to meet specialised requirements. Custom protocols are ideal for applications that are cellular but have endpoints travelling at higher speeds than supported by cellular standards. Another application is when the endpoints are all stationary but very low power consumption is needed.

A recent requirement from one of our clients was a point-to-point system in which the radio links are subject to severe shadow fading and multipath Rayleigh fading. Each link was to act as a substitute for an Ethernet wire, providing over 100 Mbps of throughput under good conditions and, more importantly, high availability low-rate communications under poor conditions.

The first step, as with any project, is to determine the requirements. Both the direct requirements specified by the client and the indirect requirements that arise from a detailed analysis and understanding of the client’s application. In the case of a protocol, the RF environment is one of the areas that need to be most carefully studied, as are the regulations that will apply to its use in different parts of the world.

An early decision was to treat the Ethernet replacement requirement literally. This meant that the user data transported by each link would consist of Ethernet frames, and the control data exchanged between endpoints would only be used for managing the link itself. Higher layer control and configuration for the client’s network would be transported as IP data and could, therefore, be secured with established security solutions. The result was that the protocol to be developed would only need to define a Physical and Medium Access Control layer.

After careful consideration, Orthogonal Frequency-Division Multiplexing (OFDM) was chosen because of its flexibility. Under good conditions, the RF channel can be packed with a large number of sub-channels, each modulated with a high order modulation, e.g. 256 QAM. Under poor conditions, the transmit power can be concentrated in just a few sub-channels and the modulation order can be dropped to 4 QAM.

Extensive simulations were carried out to confirm the suitability of the choice given OFDM’s two principal disadvantages: its susceptibility to receiver frequency errors and the need for highly linear signal chains, particularly in the transmitter. The simulations included power amplifier models and synthesizer phase noise models, as well as models of the fading environment.

The Medium Access Control layer imposed a Time Division Multiplexing structure on the transmitted signal that was designed to simplify the implementation of the client’s receivers while ensuring the system’s ability to support a range of propagation conditions. High-end throughput was sacrificed to this end, but a peak throughput of over 130 Mbps was nevertheless achieved.

Convolutional codes were used for forward error correction, and adaptive link control adjusted the number of active sub-channels, the modulation, and the code rate so that a wide range of throughputs could be supported, ensuring that the high availability requirement could be met.

It is rare that a communication system needs a custom low-level RF protocol to be developed and it is an expensive process. However, it can open up applications and services that would otherwise not be possible.

Benefits of a Proprietary Solution

Contact Us – Email: hello@plextek.com or call: +44 (0) 1799 533 200

Top