Title: Yingbo Hua
1Networking of MIMO Nodes Yingbo Hua University
of California, Riverside MURI Project Review at
UCSD July 28, 2005
2Overview - 1/4
- Networking of MIMO Nodes
- Design of MIMO Relay Nodes
- Design of Networking Protocols for MIMO Nodes
3Overview - 2/4
- Design of MIMO Relay Nodes
- 1.1 Regenerative MIMO relay nodes
(conventional) - Advantages uniform functional modules,
suitable for - large scale networking, robustness
- Disadvantages time consumed for coding,
decoding, - and handshaking for each packet at each node
-
- 1.2 Non-regenerative (NR) MIMO relay nodes
(new) - Advantages fast, virtually analog, no
handshaking - required for each packet (using freq.
division) - Disadvantages NR alone is not suitable for
many-hops - relaying (noise accumulates)
-
4Overview - 3/4
- 2. Design of Networking Protocols for MIMO
- Nodes
-
- 2.1 Route discovery algorithms
- 2.2 Data forwarding algorithms
- 2.2.1 Data forwarding without a given route
(much attention) - 2.2.2 Data forwarding with a given route
(little attention) - A given route often exists between two nodes.
- Network topology can remain relatively
stationary - with respect to a stream of packets
(w.r.t.s.p.). - But the link qualities between nodes can be
random - (w.r.t.s.p.) due to motions of the network or
changes in - the environment.
5Overview - 4/4
- We will show
- Optimal design of non-regenerative (NR) MIMO
relays - A route-guided multicast (RGMC) data forwarding
policy
6A. Optimal Design of NR MIMO Relay 1/6
Tang and Hua, IEEE SPAWC05 Tang and Hua,
IEEE T Wireless Com, submitted, July 2005
We will focus on the case
7A. Optimal Design of NR MIMO Relay 2/6
Capacity of the relay channel between source and
sink without the direct link
SNR at relay
relay-sink matrix
source-relay matrix
relay matrix
8A. Optimal Design of NR MIMO Relay 3/6
An optimization problem
Maximize the relay link capacity
subject to the power constraint at the relay
9A. Optimal Design of NR MIMO Relay 4/6
The optimal solution is
where
10A. Optimal Design of NR MIMO Relay 5/6
is monotonically increasing function and has a
unique root.
11A. Optimal Design of NR MIMO Relay 6/6
12B. A Route-Guided Multicast Data Forwarding
Policy 1/9
Ye and Hua, MILCOM 2005 Ye and Hua, IEEE T
Wireless Com, submitted, July 2005
Multicast
Unicast
13B. A Route-Guided Multicast Data Forwarding
Policy 2/9
14B. A Route-Guided Multicast Data Forwarding
Policy 3/9
- Assumptions
- Packet is received/decoded correctly iff the SINR
- exceeds a threshold
- The pdf of the received power for each packet
follows
15B. A Route-Guided Multicast Data Forwarding
Policy 4/9
A linear network
RGMC
RGUC1
RGUC2
16B. A Route-Guided Multicast Data Forwarding
Policy 5/9
ETE delay over a linear network of 100 relays
17B. A Route-Guided Multicast Data Forwarding
Policy 6/9
Progress efficiency over a linear network of 100
relays
18B. A Route-Guided Multicast Data Forwarding
Policy 7/9
We adopted the parameters of the transceivers by
Motorola Chen, ODea, Galaway, ICC02, e.g.,
Amplifier efficiency 20
Receiver noise figure 11dB
Carrier wavelength 0.125m
Transmission antenna gain 1
19B. A Route-Guided Multicast Data Forwarding
Policy 8/9
A random network
- Each node
- emits with 3
- prob.
- Each node has
- a Poisson traffic
- of average 0.16
- 225 nodes are
- uniformly
- randomly
- distributed.
20B. A Route-Guided Multicast Data Forwarding
Policy 9/9
- RGMC/RGUC ratios of efficiency, delay and
energy over a random network of 225 nodes. - Average connectivity 6.
21Conclusions and Accomplishments 1/2
- Non-regenerative MIMO relays are useful for fast
relaying - between MIMO nodes. We have discovered the
optimal - waveform shaping matrix that maximizes the
capacity of - the relay channel. Tang and Hua, SPAWC, 2005
- Route-guided data forwarding policies are
important for - large volume data transmissions over ad hoc
networks. We - have formulated and studied a route-guided
multicast data - forwarding policy. Ye and Hua, MILCOM, 2005
- A route of parallel relays with distributed
space-time coding is - an important concept. Our analysis shows that a
route of - parallel relays has a much longer lifetime than
routes of serial - relays while the increase of routing overhead is
negligible. - Ye and Hua, Asilomar, 2004
22Conclusions and Accomplishments 2/2
- List of publications partially supported by MURI
- Z. YE and Y. HUA, Networking by parallel relays
diversity, lifetime and routing overhead, The
38th Annual Asilomar Conference on Signals,
Systems and Computers, pp. 1302-1306, Pacific
Grove, CA, Nov 7-10, 2004. - Z. YE and Y. HUA, Stability of wireless relays
in mobile ad hoc networks, IEEE ICASSP2005,
Philadelphia, PA, March 2005. - X. TANG and Y. HUA, Optimal waveform design for
MIMO relays IEEE Workshop on Signal Processing
Advances for Wireless Communications, New York,
NY, June 2005. - Z. YE, and Y. HUA, "On link layer policies of
data forwarding over wireless relays", MILCOM,
Atlantic City, NJ, Oct 2005. - X. TANG and Y. HUA, "Optimal design of
non-regenerative MIMO wireless relays", IEEE
Transactions on Wireless Communications,
submitted July 2005. - Z. YE and Y. HUA, "A route-guided multicast data
forwarding policy for wireless relays," IEEE
Transactions on Wireless Communications,
submitted July 2005.
23Future Work
- Interleaving regenerative and non-regenerative
MIMO relays - should improve the network throughput. Further
investigation - will be done.
- Route-guided multicast data forwarding policy is
particularly - suitable for networks of MIMO nodes where
processing power - is not an issue. More realistic analysis and
simulation will be - done to further verify its feasibility and
performance. - Routes of parallel MIMO nodes will be
investigated. This - will achieve a new level of cross layer
cooperation of MIMO - nodes.
- Collaborations within MURI with Jim on overall
directions - JJ and Srikanth on networking protocols Mike
and Lee on NR - MIMO relays and Hamid on distributed coding for
parallel relays.