datatrak:intro

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
Next revisionBoth sides next revision
datatrak:intro [2020/05/31 11:47] – [Lane identification in Datatrak] philpemdatatrak:intro [2020/05/31 12:36] – [Three-step navigation process] philpem
Line 14: Line 14:
  
 Generally this is done by using a lower-frequency "coarse" position fix signal. The same calculations are done to obtain a position fix, but the longer wavelength increases the distance between lanes. At the coarse-acquisition frequency used in Datatrak (80Hz), the lane distance is 3747km, or 2329 miles. This is compared to the 23km (14.3 miles) lane width of the prime difference signal (13kHz difference frequency), or the 2.3km (1.4 miles) lane width of the 130kHz signal. Generally this is done by using a lower-frequency "coarse" position fix signal. The same calculations are done to obtain a position fix, but the longer wavelength increases the distance between lanes. At the coarse-acquisition frequency used in Datatrak (80Hz), the lane distance is 3747km, or 2329 miles. This is compared to the 23km (14.3 miles) lane width of the prime difference signal (13kHz difference frequency), or the 2.3km (1.4 miles) lane width of the 130kHz signal.
 +
 +
 ===== Datatrak basics ===== ===== Datatrak basics =====
  
Line 27: Line 29:
  
  
 +===== Datatrak LF signal format =====
  
 ==== Basic signal format ==== ==== Basic signal format ====
 +
 +This is the simplest possible Datatrak signal.
  
 The signal for the F1 chain (for basic Datatrak) looks like this: The signal for the F1 chain (for basic Datatrak) looks like this:
Line 34: Line 39:
 ^  $F_1$  |  Sync and timing  |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  |     |     |     |     |     |     |     |     | ^  $F_1$  |  Sync and timing  |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  |     |     |     |     |     |     |     |     |
 ^  $F_2$  |  ...              |     |     |     |     |     |     |     |     |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  | ^  $F_2$  |  ...              |     |     |     |     |     |     |     |     |  1  |  2  |  3  |  4  |  5  |  6  |  7  |  8  |
 +
 +This only allows for a single chain, and is only used for illustration.
  
 The receiver (also known as a //Locator//) tunes to $F_1$ to receive the sync and timing data and make phase measurements of the eight navigation slots against the receiver's internal temperature-compensated oscillator. The receiver then switches to $F_2$ and makes the same measurements on the F2 frequency. The receiver (also known as a //Locator//) tunes to $F_1$ to receive the sync and timing data and make phase measurements of the eight navigation slots against the receiver's internal temperature-compensated oscillator. The receiver then switches to $F_2$ and makes the same measurements on the F2 frequency.
Line 39: Line 46:
 Each navigation slot is 80ms long -- with 40ms transmitted at a higher frequency ($F_1+$ and $F_2+$) and 40ms at a lower frequency ($F_1-$ and $F_2-$). Each navigation slot is 80ms long -- with 40ms transmitted at a higher frequency ($F_1+$ and $F_2+$) and 40ms at a lower frequency ($F_1-$ and $F_2-$).
  
-=== Adding more slots: Interlacing ===+ 
 +==== Adding more slots: Interlacing ====
  
 The basic signal encoding is fine, but the limit of eight slots restricts the number of transmitters which can exist in the network. It's obvious from the timing diagram above that only half of the transmitter capacity is being used. This is easily fixed: The basic signal encoding is fine, but the limit of eight slots restricts the number of transmitters which can exist in the network. It's obvious from the timing diagram above that only half of the transmitter capacity is being used. This is easily fixed:
Line 46: Line 54:
 ^  $F_2$  |  Sync and timing  |   9  |  10  |  11  |  12  |  13  |  14  |  15  |  16  |   1  |   2  |   3  |   4  |   5  |   6  |   7  |   8  | ^  $F_2$  |  Sync and timing  |   9  |  10  |  11  |  12  |  13  |  14  |  15  |  16  |   1  |   2  |   3  |   4  |   5  |   6  |   7  |   8  |
  
-The disadvantage to this method is that the Locator can only listen to one frequency at a time -- this means that, for example, slots 1 and 9 cannot be received at the same time. This is called a //slot collision//. When planning the network, care must be taken to keep slot collisions to a minimum as Locators move around the signal area.+This allows us to have 16 transmitters, with one caveat. As the Locator only has one receiver, it can only tune to one frequency at a time. This prevents the receiver from receiving slots 1 and 9 at the same time. This is called a //slot collision//.
  
 +When planning the network, care must be taken to keep slot collisions to a minimum as Locators move around the signal area.
  
-=== Adding even more slots: Dual Cycle Interlaced mode === 
  
-Interlacing can be extended further by linking two 1.68-second //cycles// into a larger cycle:+==== Adding even more slots: Dual Cycle Interlaced mode ==== 
 + 
 +Interlacing can be extended further by linking two 1.68-second //cycles// into a larger //cycle pair//:
  
 ^  $F_1$  |  Sync and timing  |   1..8  |  9..16  |  Sync and timing  |   1..8   |  17..24  | ^  $F_1$  |  Sync and timing  |   1..8  |  9..16  |  Sync and timing  |   1..8   |  17..24  |
 ^  $F_2$  |  Sync and timing  |  9..16  |   1..8  |  Sync and timing  |  17..24  |   1..8   | ^  $F_2$  |  Sync and timing  |  9..16  |   1..8  |  Sync and timing  |  17..24  |   1..8   |
  
-This expands the system to a maximum of 24 navigation slots -- the measurements for slots 1 to 8 are updated every 1.68-second cycle, while the measurements for slots 9 to 16 and 17 to 24 are updated every two cycles (3.36 seconds).+This expands the system to a maximum of 24 navigation slots, with a further caveat: the measurements for slots 1 to 8 are updated every 1.68-second cycle, but the measurements for slots 9 to 16 and 17 to 24 are updated every two cycles (3.36 seconds).
  
-Slot collisions are still an issue, but the rules are relaxed somewhat:+Slot collisions are still an issue, but the rules change slightly:
  
   * While receiving slot $1 \leq N \leq 8$:   * While receiving slot $1 \leq N \leq 8$:
Line 69: Line 79:
     * It is not possible to receive slot $N-16$ (1..8)     * It is not possible to receive slot $N-16$ (1..8)
  
-==== Three-step navigation process ====+ 
 +===== Three-step positioning process =====
  
 (See also [[references#HoffmanWellenhoff2003|Hoffman-Wellenhoff et al, 2003]]) (See also [[references#HoffmanWellenhoff2003|Hoffman-Wellenhoff et al, 2003]])
Line 76: Line 87:
  
   * Initial "super-coarse" measurement (several-km resolution) using the $F_n+$ and $F_n-$ signals ($\Delta F = 80 Hz$)   * Initial "super-coarse" measurement (several-km resolution) using the $F_n+$ and $F_n-$ signals ($\Delta F = 80 Hz$)
-  * Secondary "coarse" measurement (several-hundred-metre resolution) using $F_2+ - F_1+$ and $F_2- - F_1-$ ($\Delta F ~= 13 kHz$) :!: use "almost" symbol +  * Secondary "coarse" measurement (several-hundred-metre resolution) using $F_2+ - F_1+$ and $F_2- - F_1-$ ($\Delta F \approx 13 kHz$) 
-  * Final "fine" measurement giving a ~50m accuracy, using the carriers themselves.+  * Final "fine" measurement giving a ~50m overall accuracy, using the carriers themselves.
  
 This system has one massive advantage: the entire navigation solution can be obtained in three stages from the same input data. This system has one massive advantage: the entire navigation solution can be obtained in three stages from the same input data.
  
-Once the position of the receiver is known, successive positioning calculations may only need the "fine" adjustment (provided it can be assumed that the vehicle does not cross a lane during successive phase measurements).+Once the position of the receiver is known, successive positioning calculations may only need the "fine" adjustment (provided it can be assumed that the vehicle has not crossed a lane during successive phase measurements).
  
 This works because a phase measurement at one frequency subtracted from a phase measurement at another frequency will result in a phase measurement taken at the difference in frequency between the two signals -- 80Hz in the case of the super-coarse fix. This works because a phase measurement at one frequency subtracted from a phase measurement at another frequency will result in a phase measurement taken at the difference in frequency between the two signals -- 80Hz in the case of the super-coarse fix.
  
-<del>At 80Hz, each lane is $L_w (metres) = \frac{1}{F} \times c$ (where $c$ is the speed of light in metres per second), or around 3747 km. The disadvantage is that the 1000 counts of phase resolution translate to a position resolution of around 3.7km per count. While this is insufficient for general positioning, it is sufficient for identifying the lane being used. Bear in mind that the use of $F1$ and $F2$ (e.g. 130kHz and 145kHz) for general positioning widens the lanes to an equivalent of 20km before the phase pattern repeats. +To put this in perspective, the relative accuracies of the different phases are: 
-</del>+ 
 +^  Phase  ^  Frequency  ^  Wavelength (lane width)  ^  Resolution  ^ 
 +|  "super-coarse"  |  $\Delta F = 80 Hz$  |  $\lambda \approx 3747 km (2329 miles)$  |  $\frac{3747 km}{1000} \approx 3.7 km$  | 
 +|  "coarse"  |  $\Delta F \approx 13 kHz$  | $\lambda \approx 23 km (14.3 miles)$ |  $\frac{23 km}{1000} \approx 23 metres$  | 
 +|  "fine"  |  $F \approx 130kHz$  |  $\lambda \approx 2.3 km (1.43 miles)$  |  $\frac{2.3 km}{1000} \approx 2.3 metres$  | 
 + 
Find me on Mastodon
  • Last modified: 2023/01/03 22:55
  • by philpem