Xiangpeng Jing

 
     
  WINLAB > Alumni > Xiangpeng Jing > My Project
   Menu
  Links
  IEEE eXplorer
IEEE Conference
IEEE CommSoc
  Contact

Xiangpeng Jing
WINLAB
NJ Tech Center II

671 Rt. 1 South
North Brunswick, NJ 08902-3390
Tel: 732-932-6857 x652 Fax: 732-932-6882
Email:

 
 
 
 
 
 
     

Projects
Bootstrapping/Discovery Evaluation using ORBIT

     The control protocols for cognitive radio networks are implemented using C++ as an network application in Linux. Specifically, the bootstrapping protocol sends periodical beacons using raw IEEE 802.11 packets. The discovery protocol calculates local link states by estimating direct link qualities from bootstrapping beacons. End-to-end path is discovered for each nodes to learn about the whole network.

     The experiments are conducted in the 20x20 ORBIT radio grid. Each node is equipped with two wireless interface, one is used for control protocols and the other is used to transmit data. Nodehandler is used to start the protocol application at each node. Perl scripts are used to collect trace files and dynamically debug each available wireless node.

Spectrum Sensor Demo System

     This demo is designed to work with the routing demo for sensor networks. In the spectrum sensor part, main work is to setup a sensor part – “spectrum sensor” and a console part with 3D ability to show the data. Both the separate two parts seem working but now it is lack of co-testing with routing algorithms.

     There are some points obtained during debugging:

 (a) The data (statistics collected when scanning channels 1 to 8) is packaged and send out to the direct associated parent after association completion.  This can be also modified to send every couples of seconds if there is a valid association with this sensor node.

 (b) There is a lagging problem for the sensed data (signal strength).  All data in different channel seems correlated and does not vary too much.  When links are down, the sensed strength can not go down very fast, and this may be due to the control of AGC in the transmitter? It is also suggested to stay longer (more than several seconds) in each channel and then sense the statistics data.

 (c) In the console, a 3D figure is drawn for the position {x,y} vs. Signal strength (dBm).  The position is actually relative position coordinate, for example, {x,y} are started from 00, ...., to 44, which has 5x5=25 values. (If there are less sensors, intermediate values can be estimated.) Default values is -80dBm and usually the signal strength for a transmitting radio is -40dBm (sensed value).

Spectrum Etiquette Protocol Implementation in WLAN-BT Scenarios  

Proof-of-concept experimental results for a CSCC implementation are presented for an example scenario in which multiple nearby 802.11b and Bluetooth devices contend for 2.4 GHz ISM band access. Large file transfer and session-based data services are tested and results show system performance based on throughput and delay is improved.

Spectrum Etiquette Demo System

The SE demo is designed to show the basic idea about spectrum etiquette protocols. The demo is setup in pairs up to 4 or 5 pairs, with 2 bluetooth pairs and 2 or 3 WLAN pairs.  The demo will be added a spectrum measurement console, which can show the dynamic power spectrum in the ISM band.  And thus interference between DS and FH radios are shown clearly.

(1)  cscc_wlan: in this part, basic CSCC protocol is implemented by IEEE 802.11 prototype. 802.11 devices are using one wireless card to broadcast CSCC messages every 2 seconds in channel #1, and exchange UDP data in another two channels (#6 and #11). UDP data is geneated as a session (~1000 packets with length ~1500Bytes each) randomly.

(2)  cscc_bt: This is for the bluetooth device with a 802.11 wireless card for CSCC (dual card). Only CSCC messages are broadcast through the 802.11 card. Bluetooth data session is also randomly generated, as a format of echo datagram (random length upto 1024 Bytes, random packet numbers). Note that Bluetooth data will consume the whole spectrum band for frequency hopping. So we can control through priority and price bid to decide wehether cscc_wlan should give up the spectrum demond to Bluetooth or vice versa.

(3)  cscc_con: This is a java console running in the same machine as cscc_bt. Because in cscc_bt, 802.11 radio do not need to switch to different channels, we can monitor the traffic pattern through this radio. This console is designed as our lab plan, and currently 802.11 traffics in ch#6 and #11, and CSCC messages in ch#1 can be monitored runtime ( with a little response delay due to the refresh interval). The console is refresed every 1 second to draw what’s happening in the whole ISM band.

Spectrum Coordination Lab Design

    - Forwarding nodes with Bluetooth USB card (Ericsson USB APtK) can communicate pear to pear.

    - Forwarding nodes with Bluetooth PCMCIA card (Xircom Creditcard) can communicate pear to pear.

Integration of Bluetooth card in forwarding node
     To support multi-devices in our sensor network, bluetooth wireless card is integrated into the forwarding node, by which small sensors with bluetooth wireless interface can access the sensor network. Data or voice can be carried through the wireless connection, and then forwarded to the bluetooth access point, or forwarded by the multi-interface forwarding node to the WLAN network.
     Technical approaches: Bluetooth PCMCIA card and USB card driver and bluetooth service stack driver are compiled as kernel modules with the linux kernel version 2.4.18. Bluez protocol stack is used and tested.
     Status: Data link (ACL) and voice link (SCO) through L2CAP layer can be setup between bluetooth devices in the forwarding nodes. Bluetooth WPAN with Access Point Gateway is being tested.
 

IEEE 802.11 Sniffer for Access Point
     The project involves micro-system using Linux kernel 2.4.18. IEEE 802.11 sniffer is implemented into the small Linux system, running in Access Points. OpenAP and libpcap packages are used for the implementation. Packets through the AP can be captured and analyzed for laboratory and measurement purposes.

  Research Highlights
 
  Cognitive Radio
   
  Spectrum Coordination

  WiMax/WiFi Network   

  WiFi/Bluetooth Co-existence at 2.4GHz    

  Spectrum Etiquette 
 
 
 
 
  Search