Ultra Band Wireless Ad Hoc Networks

1
Ultra Band Wireless Ad Hoc
Networks

• Ioannis Broustis, Srikanth Krishnamurthy
• Mart Molle
• WHYNET - UCLA, November 2004

2
Motivation

• Current wireless solutions (IEEE 802.11,
  Bluetooth) face many problems
• Limited channel capacity
• High power consumption
• UWB is an excellent solution for short-range
  communications. However, it is not intended to
  replace the above technologies, but coexist with
  them.
• 170 companies have already jumped on the UWB
  wagon

3
Why UWB?

• Advantages
• Low-power operation.
• Low cost.
• Low probability of detection and low probability
  of jamming capabilities.
• Low interference levels to existing services.
• Ability to penetrate walls, etc.
• Availability of precise location information.

4
UWB What is it?

• Any signal that
• Occupies at least 500MHz of BW, or
• More than 25 of a center frequency

5
UWB Applications

• Stream DVD content to HDTVs simultaneously.
• Wirelessly synchronize appliance clocks.
• Connect high-data rate peripherals.
• Move huge files between digital cameras,
  camcorders, and computers.
• Military applications (radars, penetrate walls..)

6
PHY Single-Band and Multi-Band

• Single-Band Implementation
• One pulse occupies the whole BW at a time.
• Multi-Band Implementation
• The 7.5GHz are divided into multiple bands.
• Information is independently encoded in the
  different bands.
• The lower limit of 500MHz, as well as the
  transmission power restrictions, must be
  maintained.
• Other possibilities OFDM, MC-CDMA (for future).

7
Why prefer Multi-Band ?

• The delay spread of the wireless channel, unless
  combated, restricts the achievable rate.
• Adaptive band selection ? Avoids interference
• Low complexity ? Small transceiver cost
• Simultaneous transmissions in the different bands

8
Time Hopping

• Time Hopping has been mostly proposed so far, to
  alleviate pulse collisions.
• Pulses are transmitted according to a Time
  Hopping Sequence (THS)
• The THS is known by both the transmitter and the
  receiver
• Receiver-based The receiver's THS is followed
• Transmitter-based The transmitter's THS is
  followed

THS1 0, 3, 2, 6
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
THS2 4, 6, 3, 3
9
Our current MAC scheme

• Advantages
• Increased channel capacity
• No need for Time Hopping
• All the multi-band advantages, discussed earlier,
  over a potential single-band implementation
• Ad hoc communications are enabled

10
Our current MAC scheme

• Increased channel capacity
• Each band is capable of achieving the same
  capacity as that of a single-band system, due to
  the delay spread effect
• No need for Time Hopping
• A band is exclusively dedicated to a pair of
  nodes
• Pulses are now being transmitted consecutively

11
Our current MAC scheme
k2
k3
k4
k5
k6
k7
12
Preliminary results

• Simulations
• Average throughput of correct bits

13
Preliminary results

• Simulations
• Frequency of collisions

14
Preliminary results

• Simulations
• Average delay for resource occupancy

15
Future Work

• MAC Layer
• Dealing with the expensive preamble period with
  UWB.
• Improving on contention resolution.
• Investigating the possibility of using OFDM/
  MC-CDMA.
• Integration with better coding schemes.
• Higher Layer Artifacts
• How does the MAC layer fit in with the routing
  layer?
• Transport layer feedback.

16
Questions? (References available upon request)
General Atomics Multi-Band Transceiver Prototype