At the PTC Congress last week in Miami, a Union Pacific (UP) panelist was asked about the real estate on top of the locomotive as to the placement of antennas in the light of the 220 MHz band that is to be used for PTC. The UP manager stated that, in fact, there can be up to 14 antennae mounted on a locomotive, and Yes! they are getting crowded. The number was quite a surprise for me in that my quick count could only come up with 10 (which is a pathetic number in itself), i.e., GEO satellite, 40MHz, 160 MHz-voice, 160 MHz-data, 450 MHz, 900 MHz, cellular, WiFi, GPS, and 220 MHz. Now, moving into the cab, the large number of antennae implies that there is a significant number of radio units mounted in nooks and crannies, with each unit most likely servicing singular applications with singular protocols via singular frequencies, e.g., voice, end-of-train (EOT), locomotive diagnostics, event recorder download, and distributed power.
Ah Yes! I can’t help but think back just 20 years ago when railroad communications engineering was so straightforward. At that time, wireless on the locomotive was limited to voice radio with the introduction of EOT as the first major use of radio telemetry across the industry. Granted, railroads were even less efficient than they are today, but there was plenty of excess of everything, so what the heck? We didn’t need wireless data to advance resource management processes. The big thrust then was the transition from crystals to the use of synthesizers to provide a full slate of channels on a single locomotive radio unit. While the railroads’ communication forces may consider their expansion of wireless data technology since then to be progressive, as suggested by the locomotive’s antenna farm, I view the transition as being totally tactical and not at all strategic. So! Is that being progressive … or is it just being evolutionary? The difference between those two perspectives is extraordinary. That is, the additional cost of being tactical instead of strategic includes an extraordinary amount of capital investment, maintenance, opportunities to delay trains due to inoperable equipment, as well as an extraordinarily poor IT architecture (both physical and logical) due to the lack of system integration resulting in an extraordinary inhibiting of advancing railroad operations in a revolutionary fashion instead of an evolutionary one. By this last point I mean that railroads have failed to use wireless technologies to advance the management of their key resources from that of being reactive to that of being proactive, as discussed in other postings on this blog.
Unfortunately, the technicians have been free to do what they like to do most, i.e., design communication solutions that are tailored to specific applications. If they were asked to justify why they didn’t consider pooling applications on a single radio, for example, they would have a number of seemingly good technical reasons, of which some would have some merit. However, the bottom line is that they have not been required to build a multi-function or multi-band wireless platform that would reduce many if not all of those extraordinary items mentioned above. This is where software define radio (SDR) comes into play.
With the term SDR being introduced as recently as 1991, it can most simply be described as replacing a number of hardware components of a radio unit with software. The underlying principle for doing so is the use of some form of digital signaling processers (DSPs) that can replace specifically designed hardware such as RF filters, mixers, amplifiers, and modulators/demodulators. While that sounds interesting, the truly great point is that a single signal processing platform can service an unlimited number of combinations of bands and protocols. It only needs the appropriate software; software which can be accessed instantaneously to provide a different radio platform to the same user.
The real breakthrough in SDR began with the rapid, exponential increase in the power of general purpose processors to service the PC market. Again, simply stated, that meant that a standard computer and the corresponding advancements in software programming could advance SDR much more rapidly than continuing to rely on the much slower advancement in specialized DSP technology. What better example of this is there than the iphone and its competitors that can handle multiple protocols, e.g., 3G, 4G, in a fashion transparent to the user?
It was a decade ago that I reached out to several companies that were then beginning to use the general purpose processors found in PCs instead of specialized DSP to deliver SDR to the military – industry complex. One such company accepted an invitation to present their concepts, using a laptop computer, to the AAR’s wireless committee. They came, they saw, and they retreated. The interest by the railroads’ technicians was one of moderate curiosity without any incentive to do anything different than what they were then doing to avoid the inevitability of meeting the FCC’s narrow-banding push of the railroads’ 160-161 MHz band. That was then, What about now? Interestingly, the answer is that the technicians have totally swung to the other extreme of the pendulum. That is, instead of spinning wheels to achieve nothing, they are totally involved in creating the ultimate wireless data network and thereby ignoring all other possibilities as to advancing technologies as well as alternative approaches to spectrum usage.
SDR is only one possibility for advancing the cost-effective and efficient use of wireless across the rail industry that will be discussed in future posts on this blog. For example, I will be addressing soon the software defined antennae (SDA), that in sync with SDR, provides the basis for cognitive radio.