Personal Statement, Ruitao Xie, Electronic Engineering PhD
I am applying for admission to Ph.D. program in *** of ***. My research interests are wireless
network and network coding in general and using network coding and optimization to analyze wireless
network in particular. I love mathematics and research with continuous digging, and I have abundant
research and engineering experience. I am confident that I am qualified for academic research in this
field.
I love research with continuous digging, and my long-term career interest is to be a excellent
professor in the research field of wireless network and network coding. My recent researches focus on
coding-aware routing in multi-rate wireless multi-hop networks, which chooses network coding pattern
among network coding, cooperative network coding and non-coding in routing selection, to improve
throughput with awareness of potential coding opportunities and link rate. In multi-rate environments,
the low rate link would decrease the broadcast rate of coding combination. Although cooperative relay
can alleviate this reduce, but the low link rate also has impact on the unicast traffic. Thus, adaptive
choice of routes with different network coding pattern which minimizes the link scheduling time slots
of path is considered in my throughput optimization formulation to calculate the maximal throughput
of unicast traffic for arbitrary topology. (这段先不看，写的内容有问题，待改)
I am digging into the other research topic, data aggregation with low delay in Wireless Sensor
Network (WSN) with single sink. My scheme is based on opportunistic routing with decentralized
fountain codes. Firstly, In WSN, the topology is dynamic since there are always some sensor nodes
failed or sleeping in any time. Thus, traditional routing protocols, which choose the best sequence of
nodes between the source and destination and forward each packet through that sequence, is not the
best approach for WSN. In contrast, the opportunistic routing, which determines each hop of a packet’s
route after transmission for that hop to use multiple transmission opportunities in intermediate nodes,
matches the opportunistic nature of WSN. In addition, when considering the energy consumption of
sensor nodes, low delay is important for data aggregation. Compared to traditional one, opportunistic
routing efficiently transmits each packet fewer times and reduces the total transmission time. In sum,
opportunistic routing could meet the requirements of low delay and dynamic topology. Secondly, the
data is only meaningful to the sink, thus intermediate nodes could make any process on the data as long
as sink could recover the original data. Decentralized Fountain Codes, which has superior decoding
complexity and randomized encoding process, is utilized at intermediate forwarders to utilize the
spatial reuse and alleviate coordination on MAC. The biggest challenge of this scheme is how to realize
the optimal degree distribution of received packets at the sink in decentralized encoding process. This
scheme is more adaptive to data aggregation in WSN than MORE in two aspects. Decoding process of
randomized network coding in MORE is computationally expensive, with a decoding complexity of
O(K3). On the other hand, this scheme could realize intra-flow and inter-flow coding, however MORE
only allows intra-flow coding whose benefit is not limited for WSN, since the data streaming in WSN
is random and burst, and the single streaming has limit packets and maintains limit time.
To become an excellent researcher in the field of wireless network and network coding, I’ve done
well in fundamental and specialized study, especially in mathematics. I had 100 in Mathematics
Analysis and Probability Theory & Stochastic Process. In order to understand communication and
network theories more deeply, I registered for course Mathematical Modeling for math majors. During
the study, I found according to Korus-Kuhn-Tucker condition and duality theorem in Linear Program,
minimum cut problem is actually the dual problem of maximum flow problem, which proves max-flow
min-cut theorem. In addition, for maximum flow problem iff there is no augmenting path in the
residual network the feasible flow shall be maximum because for Linear Program problem whose
Personal Statement, Ruitao Xie, Electronic Engineering PhD
feasible set is convex set, the local optimal solution is global optimal solution. By establishing bipartite
graph model, many practical problems, such as minimum path cover problem and bipartite graph
maximum matching problem, can be transformed to maximum flow problem to solve. As a non-math
major in this course, with flexible application of optimization theory I got awards several times in
Contest in Modeling.
Trigged by the desire to touch more advanced researches on communication and network beyond
the textbook, I opened my sight as an assistant editor of The Journal of China Universities of Posts and
Telecommunications, which gave my original opinion to do academic research.
I had accumulated abundant research and engineering experience related to wireless sensor
network and wireless optical access network. In the following summer after my junior year, I
participated in a project establishing a wireless intelligent home system based on wireless sensor
network integrating indoor environmental monitoring and health care system. I designed the vision
surveillance module of system and the routing algorithm in network layer based on ZigBee protocol. I
think that in the energy constraint network like wireless sensor network, the control overhead is critical
to reduce the energy consumption particularly in on demand routing protocol. Thus by stopping
forwarding the route discovery request when the number of transmission is beyond the length of path
found by Cluster-Tree algorithm，I decreased the number of control overhead in route discovery and
reduced the energy consumption of system.
In my first graduate year, I engaged in a project to engineer the service mobility which satisfies
seamless soft handoff and QoS of multimedia users in the wireless optical hybrid access network. I
proposed a soft handoff strategy, that gateways on Optical Line Terminal (OLT) buffer and forward
service utilizing the broadcast character of OLT, which not only effectively reduces required
equipments, but also has low packet loss and large QoS. In designing the initiation strategy of service
mobility, to increase convenience and decrease energy consumption of initiation devices, I proposed an
initiation strategy of service mobility using wireless sensor nodes, which determines handoff by
delivery probability of packets at handoff distance based on protocol model of interference. In these
periods, my research also involved with routing protection in wireless optical hybrid access network.
To improve the survivability of hybrid access network, we proposed to use the wireless link to protect
the network form fiber failure, which also increases the efficient bandwidth utilization by dynamic
allocation of bandwidth.
I appreciate for the research of Laboratory *** about ***. The other program about *** is also
attractive to me. With my academic and engineering backgrounds and research abilities, as well as my
research interest in this field, I am confident that I will be able to make a significant contribution to ***
at ***, and I hope sincerely I will have the opportunity to do so.