IEEE 802.15.4 Multipath

1
Project IEEE P802.15 Working Group for Wireless
Personal Area Networks (WPANs) Submission Title
Multipath Simulation Models for Sub-GHz PHY
Evaluation Date Submitted October
2004 Source Paul Gorday Company
Motorola Address 8000 W. Sunrise Blvd.,
Plantation, FL, 33322, USA Voice1 561 723
4047, E-Mailpaul.gorday_at_motorola.com Re
IEEE 802.15.4 Abstract This contribution
presents two multipath simulation models for use
in evaluating and comparing optional sub-GHz
PHYs. Purpose To document channel models used
for PHY evaluation. 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
Two Multipath Models

• Two analytical multipath channel models are
  defined for evaluating optional sub-GHz PHY
  performance.
• Diffuse exponential model
• Presented in 802.11 Handbook 1 and recommended
  for narrowband systems by TG3a channel modeling
  sub-committee 2
• Preferred for baseband simulations
• Discrete exponential model
• Sampled version of diffuse model
• Acceptable alternative for simulations with high
  sampling rates

3
Diffuse Exponential Model

• Diffuse each delay bin
• contains multipath energy
• Exponential average power
• decays exponentially
• Fading - each delay bin has
• independent Rayleigh fading
• Single Parameter
• RMS delay spread ? ?
• Mean excess delay ? ?
• Max excess delay (10 dB) ? 2.5?
• Max excess delay (20 dB) ? 5?

?k2 Normalized Average Power C Normalization
Constant Ts Simulation Sample Period k Bin
Number kmax ? 5?/Ts
Normalized Average Power
(Depicted ? 4Ts)
k (Bin )
Power-Delay Profile
4
Diffuse Exponential ModelBaseband Simulation

• Each channel realization can be simulated using
  an FIR filter (tapped delay line), where the tap
  weights are independent complex Gaussian random
  variables with zero mean and variance given by
  the power delay profile. In other words, the FIR
  coefficients are
• where N(m,?2) is the normal, or Gaussian, random
  variable.
• Assume quasi-static channel, such that h(k) is
  constant during packet.
• One or more packets are sent for each channel
  realization.
• At least 1000 random channel realizations for
  each PER value.

5
Diffuse Exponential ModelExample Matlab Code for
Baseband Simulation

• Variable description
• signal_in input to channel model
• signal_out output of channel model
• profile power-delay profile
• channel random realization of the channel
• kmax maximum tap index for power-delay profile
• tau RMS delay spread
• Ts simulation sampling period
• Sample Matlab Code
• kmax ceil (5tau/Ts)
• profile exp(-(0kmax)Ts/tau)
• profile profile/(sum(profile))
• channel sqrt(profile/2).(randn(size(profile))j
  randn(size(profile)))
• signal_out conv(channel,signal_in)

6
Discrete Exponential Model

• For simulations with high sampling rates, the
  diffuse model leads to long FIR filters when
  modeling large delay spreads. This increases
  complexity and reduces simulation speed.
• An acceptable alternative in such cases is to use
  a discrete, or sampled version of the diffuse
  exponential model.
• The taps (rays) are uniformly spaced by L
  samples, such that
• RMS delay spread 1.85LTs
• Max excess delay 10LTs
• Avg. power of last ray is 22 dB lower than avg.
  power of first ray

7
Discrete Exponential ModelPower-Delay Profile

• The power-delay profile for this discrete model
  is tabulated below.

k ?k2 k ?k2
0 3.95 x 10-1 6 1.97 x 10-2
1 2.40 x 10-1 7 1.19 x 10-2
2 1.45 x 10-1 8 7.24 x 10-3
3 8.82 x 10-2 9 4.39 x 10-3
4 5.35 x 10-2 10 2.66 x 10-3
5 3.24 x 10-2
8
Discrete Exponential ModelBaseband Simulation

• Each channel realization can be simulated using a
  tapped delay line, where the tap weights are
  independent complex Gaussian random variables
  with zero mean and variance given by the power
  delay profile. In other words, the tap weights
  are
• The taps are uniformly spaced by L samples.
• Assume quasi-static channel, such that h(k) is
  constant during packet.
• One or more packets are sent for each channel
  realization.
• At least 1000 random channel realizations for
  each PER value.

9
Discrete Exponential ModelExample Matlab Code
for Baseband Simulation

• Variable description
• signal_in input to channel model
• signal_out output of channel model
• profile power-delay profile
• channel random realization of the channel
• L number of samples between rays (RMS delay
  spread 1.85LTs)
• Sample Matlab Code
• profile zeros(1,10L1)
• profile(1Lend) exp(-(010)/2)
• profile profile/(sum(profile))
• channel sqrt(profile/2).(randn(size(profile))j
  randn(size(profile)))
• signal_out zeros(size(signal_in))
• for k 010
• signal_outsignal_outchannel(k1)zeros(1,kL
  ) signal_in(1length(signal_in)-kL)
• end

10
References

• 1 B. OHara and A. Petrick, IEEE 802.11
  Handbook A Designers Companion, IEEE Press,
  1999.
• 2 J. Foester, Channel Modeling Sub-committee
  Report (Final), IEEE P802.15-02/490r1-SG3a, Feb.
  2003.