A Non Coherent Receiver with Coherent Pulse Compression

1
Project IEEE P802.15 Working Group for Wireless
Personal Area Networks Submission Title A Non
Coherent Ranging Receiverwith Coherent Pulse
Compression Date Submitted June
2005 Source Gidi Kaplan, Dan Raphaeli
Company SandLinks Ltd. Address Hanehoshet 6
Tel Aviv Israel Voice, E-Mail
danr_at_eng.tau.ac.il Re Abstract
Purpose Contribution to 802.15
TG4a Notice This document has been prepared to
assist the IEEE P802.15. It is offered as a
basis for discussion and is not binding on the
contributing individual(s) or organization(s).
The material in this document is subject to
change in form and content after further study.
The contributor(s) reserve(s) the right to add,
amend or withdraw material contained
herein. Release The contributor acknowledges and
accepts that this contribution becomes the
property of IEEE and may be made publicly
available by P802.15.
2
A Non Coherent Ranging Receiverwith Coherent
Pulse Compression

• Gidi Kaplan Dani Raphaeli
• Sandlinks
• June 3rd, 05

3
Terminology

• Terminology - as agreed lately over the
  reflector
• Pulse a single UWB pulse (on the order of 1-2
  nsec)
• Burst a sequence of L UWB pulses (each pulse
  possibly modulated, the whole sequence has some
  code). Possibly, L may be between 11 to 33.
• Symbol - for data or ranging comprises of M
  bursts.
• Each pulse has energy of
• Ep Es/(LM)
• where Es is the symbol energy.

4
The Basic Idea

• Non-Coherent (NC) ranging (and demodulation) is
  considered in 802.15.4a, due to lower degree of
  complexity vs. a Coherent Receiver
• The basic NC receiver employs an energy detector
  over each UWB pulse. It suffers a considerable
  Squaring Loss due to the inherently small
  Ep/No.
• Here, we suggest a receiver which gets the full
  processing gain from each Burst of UWB pulses
  (before a square-law operation).
• It is more complex than the conventional NC
  receiver, but it allows a tradeoff of complexity
  vs. performance.

5
The basic idea (cont.)

• Assume that the receiver employs a (complex) down
  conversion to baseband in its front-end thus all
  subsequent discussion is in complex baseband.
• We further assume that over the burst, the pulses
  are bi-phase modulated, according to a pulse
  compression code.
• The latter should have a good autocorrelation
  the choice of the sequence is a different issue,
  not discussed here.

6
Correlation over the Burst

• The receiver employs a correlator over the
  (short) burst, and effectively sums up the UWB
  pulses in a coherent manner (note for ranging a
  correlator has a known delay).
• To get the symbol energy, the receiver in
  principle- sums up the squares (energies) of the
  bursts.
• Note, the receiver does not employ a phase locked
  loop to track the carrier phase
• If the burst is short enough, then even with a
  low accuracy crystal, the total phase difference
  (over the burst) is small.
• As an example, if the burst lasts over 200nsec,
  and there is 100ppm difference in freq between
  Tx. And Rx, the phase difference is less than 10
  degree for Fc4Ghz.

7
NC ranging receiver

• For ranging, the receiver has to sum up the
  energies of many bursts (taking care of the delay
  between successive bursts), in order to obtain a
  good E/N for the averaged pulse.
• It may do so by using a bank of K energy
  detectors, where each one is over D2nsec (as an
  example), and the bank covers some time window
  to account for the multipath energy spread.
• In this manner, the receiver gets an equivalent
  squared pulse, which actually has K energy level,
  one for each window of D nsec (from time 0 till
  DK nsec)
• The ranging algorithm compares the energy
  detector levels to a threshold in order to find
  the first cluster arrival time.

8
An Example

• Suppose each (ranging) symbol is composed of two
  bursts of UWB pulses, each one of relatively
  short length d1 (what we call also active time)
  followed by an interval of quiet time d2.
• The total symbol length is then 2(d1d2).
• For each burst, the receiver will employ a
  coherent correlations operation, then detect
  the energy (using a bank of d1/D energy
  detectors) of the resultant pulse
• It then has to sum up the two energy banks
  results (with the proper delays).
• This operation is done over N symbols, as
  required (by analysis or simulation) to get the
  required ranging error for a given link.
• In this case the squaring loss suffered is due to
  squaring of half the symbol (with 0.5Es/No).
• The following diagram is a (very basic) block
  diagram of the receiver.

9
An example (two bursts per symbol)