To-date, most of the Communications-Based Train Control (CBTC) systems deployed for metro lines and light rail systems have been provided and deployed by a CBTC vendor (such as Alstom, Siemens, Ansaldo STS, GE, Thales, or Bombardier) with an integrated wireless communication system, typically based either on the Wi-Fi standard or on a proprietary technology. The wireless system has not been viewed as an independent system from the CBTC system itself, although most of the leading CBTC vendors claim that their systems are agnostic as far as the wireless communication system is concerned. As the Internet of Things paradigm develops further in the mass transit space, I expect the wireless communication portion of any CBTC system to become more and more independent from the actual train control system. The wireless network connecting the train to the wired network and to the Internet will be designed and deployed as a reliable and independent infrastructure where multiple types of traffic will co-exist with different quality of service (QoS) levels.
By reviewing the specifications of several CBTC systems and their communication components, I’ve noticed a significant concern of the network designers is joining critical CBTC-related traffic together with non-critical traffic such as video streaming and data streaming. However, the desire to keep CBTC traffic on a totally independent system results in network designs that demand a large number of radios devices. These can create a lot of interference because they’re co-located on the train and are fairly close to each other along the track way-side. Although the radios for the CBTC and non-critical applications can be operated on different channels, the noise floor in the area increases due to multiple transmissions on channels that are fairly close to each other in the spectrum. A higher noise floor results in interference and a higher error rate during transmission. Two physical separate systems also mean higher installation and maintenance costs.
On the other hand, correctly designing a fully redundant system running on multiple channels for spectrum redundancy – but not using multiple nearby channels simultaneously – can improve the overall network performance without impacting reliability and redundancy. Moreover, the increased reliability with less deployed hardware will reduce downtimes and costs at the same time. Properly implemented Quality of Service (QoS) and traffic engineering policies can facilitate the critical, traffic guaranteed throughput and minimal latency, increasing the availability of the wireless network.
Our approach to designing train to ground wireless communications networks, that can, at the same time, be used for critical CBTC traffic and for non-critical video and data streaming, involves a redundant architecture:
– Deploying base stations along the track of the train that guarantee two base stations in clear line of sight (LOS) with the train at all times
– Alternating carrier frequencies of the base station to create spectral redundancy and increase the reliability of the system against denial of service attacks (DoS)
– Redundant power inputs for all wireless devices to avoid communication failures resulting from a power failure
– Redundant mobile wireless devices on board each train, operating on at least two frequencies at the same time
– Employing MIMO-based radios technology and dual-slant dual-polarity antennas to increase signal-to-noise ratio (SNR) and leverage the multi-path in tunnels to increase throughput and distance between base stations
– Advanced traffic engineering leveraging an MPLS infrastructure to guarantee throughput and latency to critical traffic CBTC in all network conditions
Fluidmesh architectures and technology have been proven to provide up to three times the throughput and allow up to twice the distance between way-side base stations compared to other wireless technologies based on the 802.11 standard. The improvement in performance is achieved with multiple technologies working together such as MIMO chipsets and dual-polarity antennas, MPLS label switching and traffic engineering, and adaptive rate control algorithm based on artificial intelligence (learning)
Multiple-input and multiple-output (MIMO) wireless technology has not been widely applied to CBTC and train communication to date. However, Fluidmesh deployments in metro tunnels have proven that MIMO, in conjunction with the right type of antenna technology, can leverage the multiple path that the wireless signal creates by bouncing on the walls and ceiling of the tunnel, creating multiple uncorrelated transmissions that improve SNR and throughput. While the performance of most wireless systems is negatively affected in a tunnel environment, MIMO-based Fluidmesh products deliver in real metro tunnels up to 100 Mbit/sec of usable throughput spacing base stations twice as much compared to a Single Input Single Output (SISO) legacy wireless system.
In the coming years, the advancements in wireless technology and the significant increase of throughput requirements for trains and mass transit vehicles will drive demand for broadband train to ground communication systems capable of providing a low bandwidth, highly reliable wireless link for CBTC and other vital applications. There will also be more demand for a broadband link to stream live video and provide on-board Wi-Fi for passengers. In the coming years I anticipate seeing more advanced network designs with a unique, carrier-grade, highly available and broadband wireless network connecting the train to the way-side fiber network, leveraging the state of the art in terms of wireless technology in tunnels.
CEO and co-founder