Using Waveguides in Light Beam Traffic Systems

What good is a new discovery, if you can`t celebrate it with firecrackers ?

Anfang ne proposed solution to the task of ensuring good communications with the beamcars is to use waveguides in the ceiling inside the beams. These waveguides would have a slit underneath, just like the beam itself, where the electromagnetic wave would leake out to be captured by the beamcar´s antenna. Likewise, the beamcar´s transmitter would, through this slit, send its information into the waveguide.
This page provides som basic waveguide theory and then investigates in detail how this would work under all conceivable situations. We will also point at some problems with this alternative.
It should be noted that the idea of using waveguides originated before the Bluetooth communication protocol saw the light of day. Using Bluetooth, waveguides are not needed for ensuring 2-way communication with the beam vehicles.
It could be, though, that the waveguides might serve:
A) As a backup communication system
B) To provide other kinds of communication, for the passengers benefit.
The general idea is to use the hollowness of the beams for some mode of radiowave-carried communication between stationary antennas and the beam vehicles. In this way, we can set our own terms for this communication, without causing radiowave disturbances in the surroundings, and also without having to adhere to any regulations regarding radiowave transmissions. Inside the beams, we are free to solve this problem in whatever manner we choose.
Waveguides have the advantage of providing reliable communication which is not hampered by reflections and noise inside the beams. But there are communication protocol issues that have to be solved.

Some waveguide theory
The physical makeup of waveguides
Mounting the waveguides in the beams
Antennas
How the waveguides would be used

1. Some Waveguide Theory

aveguides are used for transferring signals where the wavelengths involved are so short that they are of the same size as the electronic circuits. Electronic currents create magnetic fields that radiate away from the currents. These magnetic fields "steal" energy from the current. At low frequencies, this energy is mainly re-absorbed by the current, as the direction of current changes. As the frequency is increased, however, proportionately more of the energy is lost to this magnetic field, which in turn creates an electric field. Figure 1:1 At sufficiently high frequencies, most of the energy at any given moment is stored in this combined electro-magnetic field, and not so much in the electric current itself. At this point, it makes more sence to transmit the energy through this electro-magnetic field, rather than in the electric current in the wire. Waveguides can have a circular or a rectangular cross-section. They can have other shapes as well, if motivated for some reason or other. They need not even be three-dimensional. In the light beam system-context, we will only concern ourselves with rectangular waveguides. Depending on their shapes and dimensions, they can transmit certain frequencies in what is termed modes. These modes are also determined by the fact that waveguides are made from electrically conducting material, often aluminum. Thus, the electric component of the electro-magnetic field at the walls of guides are always zero. Four conducting modes are shown very simplified in illustration 1:2. We have illustrated the electro-static field lines in red and the electro-magnetic field lines in blue. The digits in the subscripts for each mode describe the characteristica for the electric and magnetic field components of the mode in question. As this is no ground course in waveguide theory, we have to refer the interested reader to other websites or to a good book about waveguide behavior. We will limit the discussion below to only concern the most basic transmission mode, the so-called TE₁₀ -mode. Different modes can be used simultaneously in the same waveguide, but let´s not, for the moment, unnecessarily complicate things.	Figure 1:2

aveguides are used for transferring signals where the wavelengths involved are so short that they are of the same size as the electronic circuits. Electronic currents create magnetic fields that radiate away from the currents. These magnetic fields "steal" energy from the current. At low frequencies, this energy is mainly re-absorbed by the current, as the direction of current changes. As the frequency is increased, however, proportionately more of the energy is lost to this magnetic field, which in turn creates an electric field.

Simple drawing of magnetic field around an electric conductur

Figure 1:1

At sufficiently high frequencies, most of the energy at any given moment is stored in this combined electro-magnetic field, and not so much in the electric current itself. At this point, it makes more sence to transmit the energy through this electro-magnetic field, rather than in the electric current in the wire.

Waveguides can have a circular or a rectangular cross-section. They can have other shapes as well, if motivated for some reason or other. They need not even be three-dimensional. In the light beam system-context, we will only concern ourselves with rectangular waveguides. Depending on their shapes and dimensions, they can transmit certain frequencies in what is termed modes. These modes are also determined by the fact that waveguides are made from electrically conducting material, often aluminum. Thus, the electric component of the electro-magnetic field at the walls of guides are always zero.

Four conducting modes are shown very simplified in illustration 1:2. We have illustrated the electro-static field lines in red and the electro-magnetic field lines in blue.

The digits in the subscripts for each mode describe the characteristica for the electric and magnetic field components of the mode in question. As this is no ground course in waveguide theory, we have to refer the interested reader to other websites or to a good book about waveguide behavior. We will limit the discussion below to only concern the most basic transmission mode, the so-called TE₁₀ -mode. Different modes can be used simultaneously in the same waveguide, but let´s not, for the moment, unnecessarily complicate things.

Examples of waveguide electro-magnetic field patterns

Figure 1:2

2. The Physical Makeup of Waveguides

he proposed waveguide would have a rectangular profile as shown in figure 2:1. For this mode, there is a cutoff-frequency, f_c, which depends on the width a of the waveguide, according to the formula *f_c = c / 2 a where c is the speed of light ( = 300 000 km/s). Thus, if we for instance put a = 10 centimeters, we would get f_c = 300 000 000 / 2 * 0.01 = 15 GigaHertz. If we wanted to double that frequency, we could either use the TE₂₀ - mode, or we could make the waveguide half as wide, i.e. put a = 5 centimeters. All frequencies below f_c will be attenuated, while frequencies that are higher than f_c** will be propagated.	One solution that would allow beamcars and nodes to use full duplex (i.e. transmit and receive simultaneously) would be to use at least 2 carrier wave frequencies in each beam segment. One wave would be generated by the stationary antenna inside the beam, belonging to the node computer. The other carrier wave would be generated by the beamcar. If there are more than one beamcar on a particular waveguide segment, they would have to pick transmission frequencies from a pool of allowavable (and free) frequencies. Another solution would be for all parties on a particular waveguide segment to use one and the same carrier frequency. Beamcars and nodes would all use collision detection to avoid transmitting simultaneously. This solution would be simpler, and still probably quite sufficient, because the waveguide segments would probably not be longer than about 600 meters, partly because of attenuation of the signal (since some of it would disappear through leakage) and partly because the beam network we are discussing would have nodes at frequent intervals.	Figure 2:1

3. Mounting the Waveguides in the Beams

Principle of waveguide segments, shown from one side

Figure 3:1

Principle of waveguide segments, shown from above

Figure 3:2

The waveguides would be mounted in the ceiling of the beams. Because:

The signal from the beamcars as well as from the stationary transmitters get somewhat attenuated along the waveguides
We need to limit the amount of signalling traffic so as to avoid an unnecessarily high proportion of call collisions

he waveguides on long continous stretches of beams whould have to be broken down into segments of suitable lengths. For attenuation reasons, this length should not be longer than about 600 meters. Figure 3:1 illustrate this principle; albeit it is greatly out of proportion; it just shows the idea. The waveguide segments would have vertical walls at each end, resulting in small gaps between the segments.

The gaps between these segments would thus not be physically big, but they mean a potential complication if the beamcar is in the middle of a transmission/reception when crossing these gaps.

In order to avoid unnecessary complications (such as waves from one transmitter spreading into different beams), the waveguides would also be separated into segments at each beamshunt, as shown in figure 3:2. The gap indicated by "1" would be larger than the one at "2". The node antenna for both transmission and reception would be placed at the middle of each segment. Or, it might work better if the node computer had its stationary transmitting antenna, at one end of each segment, and at the other end its receiving antenna.

The 2 antennas of the propulsion vehicle

Figure 3:3.

Underneath, the waveguide would have a slit, where the beamcar´s receiving antenna can catch the signal as it leaks out through the slit, as shown in figure 3:3. The beamcar also has a transmitting antenna which directs the signal through the slit into the waveguide.

The gap between the waveguide segments could theoretically be covered by the leaking fields, and if the 2 segments belong to the same node, which uses the same transmitting frequency in both beams, the communication could possibly be carried on without interruption when the beamcar passes a shunt in the beam and its antennas thus also pass through a gap between waveguides. As a complement, one could also use a communication protocol that allows for re-transmission of lost data blocks.

Yellow indicates the leaking electro-magnetc field from the waveguide
Figure 3:5.
Orange indicates the leaking electro-magnetc field overlap from two waveguide segments
Figure 3:6.

Parallel waveguides and backup-antennas
Figure 3:7.
Anfang et´s look a little closer as to how this works. In this example we will use the TE₁₀ - mode mentioned above. That means that all antennas would be horizontally positioned, at right angles to the waveguides, and almost as long as the waveguide is wide, as shown in figure 3:7. For higher carrier-wave frequencies, they would be correspondingly shorter.
Through a slit at the bottom of the waveguide, the electrostatic field energy would leak out and reach the beamcar antennas travelling below. The shortend view in figure 3:5 is shown enlarged in figure 3:7, where the yellow depicts the area where the field is strong enough to be usable. Figure 3:6 shows the end-to-end overlap between waveguide segments that would occur at the shunts, because there is some sideways spreading of the signal. This overlap might not be long enough in duration to have a reliable functionality, wich means that the receiving equipment might not be able to piece together dismembered data blocks that have been received by different antennas. In other words, the overlap should be long enough in duration to allow complete transmission/reception of the longest data block by at least one of the waveguide segments. Phrased in another way, if the longest possible block of data is T seconds in duration, the passage of the beamcar antennas through the overlap should be at least T seconds in duration. The speedier the beamcar, and the longer the data blocks, the longer the overlap needs to be.
If the end-to-end signal overlap is not long enough, one way to solve this problem would be to have parallel waveguides through the shunts, as shown in figure 3:8.
Parallel waveguides to eliminate signal interruption as beamcar passes a gap
Figure 3:8.
Two alternative arrangements of parallel waveguides
Figure 3:9.
To handle this situation, the beamcar would need to be equipped with at least 2 (maybe 3) antennas next to each other. The boxes marked "B" in the illustration figure 3:9 indicates buffering of the signal, after demodulation and decoding. The on-board computer on the car must check all 3 buffering areas continually and be able to separate useful information from garbage.
One way to avoid several antennas and signal buffering areas on the propulsion car would be to tilt the outer waveguides inwards, as shown in B in figure 3:9. That is probably a simpler solution than alternative A.

The general idea is that the node will only communicate with a specific beamcar by using the one particular waveguide segment that is used by that beamcar at the moment. Keeping track of all beamcars would thus be a requirement.
Now, in the point-synchronous traffic system all cars approaching a weaving shunt would have passed a booking point, and thus been alloted a timeslot by the node. Assuming right-side traffic in figure 3:10, the node will know that cars on waveguide segments 4 and 5 will be on segment 6 after a certain time. Likewise, cars on segments 6 and 7 will soon be on segment 8 and cars on segments 3:6 and 3:7 will soon be on segment 11. Leaving a berth, as a beamcar on segment 10 would be doing, is treated the same as passing a booking point; the car would be alloted a timeslot. Cars that get stopped for some reason or another, would announce this to the node. The node would note that fact, and also note which segment the car is transmitting from. The node would then (of course) use that segment when communicating with the beamcar.
Then there are the divergence shunts. A car wishing to berth would of course announce this, and the node would assume that a car on for instance segment 8 will soon be on segment 10. For cars passing through, there could be 3 scenarios:
Illustrating beamcar control at a station
Figure 3:10.

The car announces each time which way it will switch. This would generate a lot of communication.
The car receives its travel instructions through the whole synchronous area from the node.
The car announces its destination when passing through the first booking point. If nodes and cars are using the same routing tables, the node would then know how the car will travel through the area.

4. Antennas

he transmission and reception characteristics of a dipole-antenna is shown in figure 4:1. The length of the red arrows indicate the signal strength in the direction that the arrows are pointing. These characteristics can be greatly modified by various means, if need be. The lower part of figure 4:1 shows how this antenna would be placed in the middle of each waveguide segments, radiating and receiving from both directions. Ordinary transmission cables (or fiber-optic cables) would connect these antennas to the node computer, as shown in figure 4:2. The beam cars would probably be best served by using directional horn antennas, as shown in figure 4:3. Attenuation The transmission node that suffers least from attenuation is the so-called TE₁₀. For most metals (such as aluminum), the attenuation is below 0.02 dB/meter for frequencies between 5 and 20 GigaHertz. A power loss of half corresponds to about 3 dB, which would be about 150 metes. A power loss of 90% would correspond to about 475 meters. But these figures apply to waveguides that do not leak. Our waveguide will have a slit along its bottom, and the ensuing leakage will have to be calculated, using various relevant parameters.	Figure 4:1 Figure 4:2 Figure 4:3

he transmission and reception characteristics of a dipole-antenna is shown in figure 4:1. The length of the red arrows indicate the signal strength in the direction that the arrows are pointing. These characteristics can be greatly modified by various means, if need be. The lower part of figure 4:1 shows how this antenna would be placed in the middle of each waveguide segments, radiating and receiving from both directions. Ordinary transmission cables (or fiber-optic cables) would connect these antennas to the node computer, as shown in figure 4:2.

The beam cars would probably be best served by using directional horn antennas, as shown in figure 4:3.

Attenuation

The transmission node that suffers least from attenuation is the so-called TE₁₀. For most metals (such as aluminum), the attenuation is below 0.02 dB/meter for frequencies between 5 and 20 GigaHertz. A power loss of half corresponds to about 3 dB, which would be about 150 metes. A power loss of 90% would correspond to about 475 meters.

But these figures apply to waveguides that do not leak. Our waveguide will have a slit along its bottom, and the ensuing leakage will have to be calculated, using various relevant parameters.

Figure 4:1

Interconnecting stationary antennas belonging to a node or regional computer

Figure 4:2

Beam vehicle´s use of directional antennas for both transmission and reception

Figure 4:3

5. How the Waveguides would be used

e will here look at situations that could arise, and see whether the beams can handle them all in a satisfactory manner. There would be 3 kinds of communication, using the waveguides:

To control traffic flow.
To provide travel-related information to travelers.
Possibly also to provide continuous communication of an entertainment nature to travelers.

Waveguides are not needed to provide travel-related information or broadband communication of an entertainment nature. There are other methods to provide this service, such as radiating antennas inside the beams. Which method is best for this purpose remains to decide. Below, we will look more closely at the first (and most important) function of the waveguides: To communicate with and to control the beam vehicles.

1. To Control Traffic Flow

Assumption 1: On beams with asynchronous traffic, the cars would be communicating directly with the Central Computer. On beams with synchronous traffic, the cars would be communicating with the local node which is in charge of this particular beam segment. As a rule, it would be the node that the car is approaching.

Assumption 2: All beam segments could use the same transmission frequency, since these segments are isolated from each other, and there would only be one stationary transmitter/receiver in each segment (albeit possibly having several antennas distributed in the waveguide segment). I.e. the stationary antennas all use the same carrier wave for transmission. That simplifies this system, as compared to a cellular phone-system.

Assumption 3: All cars would also use the same transmission frequency, and they would use the same frequency the nodes are using. When several cars want to transmit simultaneously in the same waveguide segment, a collision detection scheme would have to be resorted to, where both (or all) cars disband transmission and try to retransmit a few milliseconds later. However, since the node usually has most to say, too much call collisions should lead to a decision to let the fleet of cars use a different carrier frequency from that of the node.

Motivation: Letting each car have its own individual transmission frequency is not viable in a network that might grow to several thousand cars. Letting each car have its own transmission frequency within a segment, taken from a pool of available and free frequencies for that segment would be difficult to administer. It would also generate extra communication.

Consequence 1: A typical, centrally located node in a big network would have quite a lot of beam segments to manage. In each segment there would be a transmission and a receiving antenna. There could be more than a hundred vehicles for the node to manage at any one moment. Information needs to be exchanged when:

An approaching car passes a booking point and gets assigned a timeslot
An approaching car asks for a berth (any berth or a particular one)
An approaching car gets queued or assigned to a free berth by the node
A landing car announces that it is occupying a berth
A leaving car announces that it makes a berth available
A leaving car tells the node of its destination and asks for a timeslot
A leaving car gets assigned a timeslot by the node
A car reports some kind of problem
The node reports some kind of problem
A buffered car gets an assignment to go somewhere

All this means that the node knows at all times in which beam segment a specific car is. The node computer has to know this, or this communication scheme would break down. The node (or regional) computer could thus filter its outgoing traffic so that messages only get transmitted to the waveguide segments where the car in question is likely to be.

This reduces the task of listening for messages for all the cars. But the real gain is that call collisions will only occur between cars that happen to be on the same segment, and not between all the cars within this node´s jurisdiction.

How well would this arrangement function for a travelling car that passes a gap between two segments during on-going information exchange? Well, as stated above, one could use a communications protocoll that allows for re-transmission.

How well would it function for a travelling car that leaves one node area and arrives in another? Should not be any problem there either, since the first node knows that the car is leaving its area (from reports from a sensor, for instance), and the second node will be informed of the approaching car when it passes a booking point.

The question of protocol

The reader will note that this page has only concerned itself with how waveguides would function as a transmission medium. We have not here considered what communication protocol to be used. Call collision, garbled messages, interruptions in communications, due to the vehicle passing from one waveguide segment to another; all these require retransmissions. Time can be crucial when travelling at high speed, since received messages also have to processed and acted upon, so this is not satisfactory.

Great demands would thus have to be made on such a communication protocol. And one protocol that fill all these demands now exist: Bluetooth. Bluetooth could possibly be made to accomodate the waveguides. But Bluetooth does not need these waveguides in order to function.

The question that remains to be solved is: Would Bluetooth benefit from using waveguides, or would it be hampered by using them? Practical tests would probably have to be made to answer this question. If FlyWay could do without the waveguides, beams could be manufactured to a somewhat lower cost.

Copyright © 2004, SwedeTrack System.	Last Updated: 2007-01-17		This site is maintained by Johnson Consulting