Network Coding in Disruption Tolerant Network

Presented by- Faheema Monica (S20141501) University of Science and Technology, Beijing

Contents

Fundamentals of Disruption

Tolerant Network

• Introduction to DTN• Working principles• Bundle protocol• DTN vs Classical networking• Potential applications of DTN• DTN’s applicability to NASA

Applications of Network Coding

in DTN

• Random Linear Coding• Coding benefits for Broadcast communication• Coding benefits for Unicast applications• Network Coding Improves Delay vs. Transmission Number Trade-

off• Conclusion

DTN (Disruption Tolerant Network)

A networking architecture provides communication in unstable and stressed environments where the network would face disruptions, high bit error.

DTN provides: reliable data transfer without complete cotemporaneous end

to end path to destination, retransmission from closet relay node rather than sender, custody transfer and return receipt operations, DTN can reduce delay and increase throughput.

Store-Carry-Forward DTN routing adopts a so called store-carry-forward paradigm. Under this paradigm, each node in the network stores a packet that has been forwarded to it by another node, carries the packet while moves around, and forward it to other relay nodes or to the destination node when they come within transmission range.

Working principles of DTN

Bundle ProtocolA different approach to TCP/IP: BP ;RFC 5050

DTN vs. Classical(Wireless) Networking

DTN

Classical networking

DTN vs. Classical Networking

Classical networking

DTN

DTN vs. Classical Networking

DTN

Classical networking

DTN vs. Classical Networking

DTN

Classical networking

Potential applications of DTN:

1. Space Agencies: 2. Military and Intelligence:3. Commercial: 4. Public Service and Safety: 5. Personal Use: 6. Environmental Monitoring:

Space operations require coordination of data transfers across the space link with the scheduled ground contacts. This is done manually by DTN.

DTN enables the decoupling of these two activities by allowing onboard data to be held in persistent storage until a contact is available, at which time it is automatically sent to the Command and Data Handling (C&DH) for downlink.

DTN’s applicability to NASA and International Space Mission

Random Linear Network coding is a technique, which can be used to improve a network's throughput, efficiency and scalability, . Instead of simply relaying the packets of information they receive, the nodes of a network take several packets and combine them together for transmission. This can be used to attain the maximum possible information flow in a network. In broadcast transmission schemes allows close to optimal throughput using a decentralized algorithm. Nodes transmit random linear combinations of the packets they receive, with coefficients chosen from a Galois field. If the field size is enough large, the probability that the receiver(s) will obtain linearly independent combinations (and therefore obtain innovative information) approaches 1. It should however be noted that, although random network coding has excellent throughput performance, if a receiver obtains an insufficient (or lost) number of packets, it is extremely unlikely that they can recover any of the original packets. This can be addressed by sending additional random linear combinations until the receiver obtains the appropriate number of packets.

RANDOM LINEAR NETWORK CODING

Let see the Example of the butterfly network coding:The butterfly network is often used to illustrate how random linear network coding can outperform routing. Two source nodes have information A and B that must be transmitted to the two destination nodes (at the bottom), which each want to know both A and B. Each edge can carry only a single value (we can think of an edge transmitting a bit in each time slot).

If only routing were allowed, then the central link would be only able to carry A or B, but not both. Suppose we send A through the center; then the left destination would receive A twice and not know B at all. Sending B poses a similar problem for the right destination. We say that routing is insufficient because no routing scheme can transmit both A and B simultaneously to both destinations. Using a simple code, as shown (random linear network coding), A and B can be transmitted to both destinations simultaneously by sending the sum of the symbols through the center – in other words, we encode A and B using the formula "A+B". The left destination receives A and A + B, and can calculate B by subtracting the two values. Similarly, the right destination will receive B and A + B, and will also be able to determine both A and B.

Theorem 1:

We assume that all packets are of the same length with P bits payload. When RLC is used in packet data networks, the payload of each packet can be viewed as a vector over a finite field, Fq of size q, more specifically, where the addition and multiplication operations are over Fq .

The coefficient =( 1,... k) is called the encoding vector, and the α α αresulting linear combination, x is an encoded message. We say that two or more encoded messages are linearly independent if their encoding vectors are linearly independent. Each original packet, mi, can be viewed as a special combination with coefficients i = 1, and j α α= 0, j ≠i.∀

 

 

Control Signaling

Before presenting the benefits of RLC in DTN, let have a look to the different existing control signaling:

Control Signaling Because of the ad hoc nature and dynamically changing topology of DTNs, nodes perform

beaconing in order to discover their neighbors (via broadcasting periodic beacon packets), and/or exchange with neighbors information about packets/coded packets they carried. Such control signaling is useful for nodes to decide whether to transmit and what information to transmit. The following different levels of control signaling have been considered: 

No Signaling: Under this most basic case no information about the neighborhood is available. Nodes decide to transmit packets without knowing whether there is a neighboring node or not.

Normal Signaling: Under normal signaling, each node periodically transmits beacon messages in order to discover neighboring node, nodes within its transmission range. With normal signaling, a node typically only transmits information when it detects at least one neighbor.

Full Signaling: Under full signaling, each node not only performs periodic beaconing to discover its neighbors, but also exchanges with its neighbors information about what packets or coded packets are stored locally, the sequence numbers of packets or the encoding vectors of coded packets. Based on such information, a network node typically only transmits to its neighbors if it has useful information for them.  

Coding Benefits for Energy Efficiency:Consider a network of N nodes, where each node has generated a packet to be broadcast to all the other nodes. Assume nodes move according to the uniform at random mobility model, i.e., at each time slot each node independently jumps to a new location in the terrain selected uniformly at random. At each time slot, each node decides to turn off or on its radio respectively, with probability p and 1-p. Assume there is no control signaling (no information about neighboring nodes and the information they carry). In each time slot, each node that is turned on randomly chooses a packet to transmit (under a non-coding scheme), or transmits a random linear combination of its coded packets to its neighbors (under an RLC scheme). There are on average (1-p) N transmissions in the network at each time slot.

Theorem 2: Broadcasting to all receivers can be achieved using on average the:

time slots, without using network coding;

time slots, using network coding with a large enough field size, q.

Which one is better?  Of course Tnc is better with a small time slot than Tw.

Coding Benefits for Broadcast Network Communication

Simulation results have confirmed that the block delivery delay under the RLC scheme is very close to the minimum block delivery delay, as shown in FIG5-A, which plots the empirical cumulative distribution function (CDF) of minimum block delivery delay, and the block delivery delay achieved by the RLC and the non-coding scheme over 100 different simulation runs each

with a different random seed.

Coding Benefits for Unicast Network Communication

Figure 10.5: DTN with N=101 nodes, homogeneous exponential inter-meeting time with rate"=0.0049, bandwidth constraint of b=1 packet per contact, and unlimited buffer space.

The RLC scheme improves the delay versus number of transmission trade-offsSimulation studies reported compared the block delivery delay versus transmission number trade-off achieved by the non-coding scheme with binary spray-and-wait applied to each of the K packets, the token- based RLC scheme and the E-NCP scheme. Figure 10.7 plots the average block delivery delay versus number of transmissions, for a block of K=10 packets, under different token limits, for the cases both without buffer constraints (a) and with buffer constraint of B = 2 (b). We observe that, with a similar number of transmissions, the RLC schemes achieve smaller block delivery delay than non-coding schemes, and the token-based RLC scheme outperforms the E-NCP scheme, especially for small numbers of transmissions. The results for a limited relay buffer case further establish the benefits of the RLC schemes in reducing block delivery delay without increasing transmission overheads：

Network Coding Improves Delay vs. Transmission Number Trade-off

Figure 10.7: Block delivery delay vs transmission number trade-off under the same network setting as Fig. 10.5except for the bandwidth and buffer constraints.

Figure 10.9: Block delivery delay vs transmission number trade-off with full signaling and normal signaling, the network setting is the same as that of fig 10.5

RLC and no coding scheme under different control signaling:

In this presentation, at first we have introduced the DTN concept, presented the opportunities offered by DTN, and shown how DTN-based communication may represent an opportunity for satellite networking.

And then in later part, our main focus was to show the performance evaluation of RLC based routing schemes both for broadcast and unicast network communications, unicast network communications and broadcast network communication having been the object of a larger amount of research. We highlighted both theoretic results and simulation studies findings. For both communication models, the RLC based scheme provides better trade-off between energy consumption and delivery performance.

Conclusion