# The Capacity of Wireless Networks

## Presentation on theme: "The Capacity of Wireless Networks"— Presentation transcript:

The Capacity of Wireless Networks
Piyush Gupta and P. R. Kumar Presented by Zhoujia Mao

Outline Arbitrary networks Two models: protocol and physical
An upper bound on transport capacity Constructive lower bound on transport capacity Random networks Constructive lower bound on throughput capacity Conclusions

Arbitrary Networks n nodes are arbitrary located in a unit area disc
Each node can transmit at W bits/sec over the channel Destination is arbitrary Rate is arbitrary Transmission range is arbitrary Omni directional antenna When does a transmission received successfully ? Allowing for two possible models for successful reception over one hop: The protocol model and the Physical model We will consider two types of networks: arbitrary networks and random networks , where the nodes and their destinations are randomly chosen Arbitrary network is defined as … The reason to look at those networks is to not to be confined with any parameters of the model, and thus to find upper and lower bounds that suitable to any network model. We will result in an upper bound, meaning that even under optimal circumstances, the throughput will never exceed this bound We will look at those two models for successful reception of a packet: protocol model and physical model.

Protocol Model Let Xi denote the location of a node
A transmission is successfully received by Xj if: For every other node Xk simultaneously transmitting D is the guarding zone specified by the protocol

Physical Model Let Be a subset of nodes simultaneously transmitting
Let Pk be the power level chosen at node Xk Transmission from node Xi is successfully received at node Xj if: Models a situation where minimum SIR is necessary for successful receptions Signal decays with distance as 1/r^alpha

Transport Capacity of Arbitrary Networks
Network transport one bit-meter when one bit transported one meter toward its destination Main result : Under the Protocol Model the transport capacity is ( as n ) if nodes are optimally placed, the traffic pattern is optimally chosen and the range of each transmission is optimally chosen Transport capacity is measured with bit meters.

Arbitrary Network – upper bound on transport capacity
Assumptions: There are n nodes arbitrarily located in a disk of unit area on the plane The network transport lnT bits over T seconds The average distance between source and destination of a bit is L Transport capacity of lambda*n*L is achieved In the time slot each node transmitting according to its arbitrary rate The slotting assumption can be eliminated.

Theorem 2.1 In the protocol model, the transport capacity lnL is bounded as follows: In the physical model, We will prove only the protocol model bound.

Remarks The upper bound in Protocol Model only depends on dispersion in the neighborhood of the receiver The upper bound in Physical Model improves when α is large, i.e., when the signal power decays more rapidly with distance When the domain is of A squares meters rather than 1 m^2, then all the upper bounds above are scaled by

Arbitrary Network – constructive lower bound
Theorem 3.1: There is a placement of nodes and an assignment of traffic patterns such that the network can achieve under protocol model Proof – define r := Place transmitters at locations: Place receivers at locations:

A constructive lower bound on capacity of arbitrary network
Dr >(1+D)r (( )) (( )) (( )) r 2Dr (( ))

Random Networks n nodes are randomly located on S2 (the surface of a sphere of area 1sq m) or in a disk of area 1sq m in the plane Each node has randomly chosen destination to send l(n) bits/sec All transmissions employ the same nominal range or power Two models: Protocol and Physical S^2 introduced here to separate edge effects from other phenomena

Protocol Model Let Xi denote the location of a node and r the common range A transmission is successfully received by Xj if: For every other Xk simultaneously transmitting Very similar to the arbitrary case, but here we have got a constant r

Physical Model Let Be a subset of nodes simultaneously transmitting
Let P be the common power level Transmission from node Xi is successfully received at node Xj if: Here there is only one power level: P common to everybody

Throughput Capacity of Random Networks
Feasible throughput: λ(n) bits per second is feasible if there is a spatial and temporal scheme for scheduling transmissions such that every node can send λ(n) bits per second on average to its chosen destination Throughput capacity: throughput capacity of the class of random network is of order θ(f(n)) bits per second if there are constants c > 0, c’ < ∞ such that

Spatial tessellation Let {a1,a2,….ap} be a set of p points on S2
The Voronoi cell V(ai) is the set of all points which are closer to ai than any of the other aj’s i.e.: Point ai is called the generator of the Voronoi cell V(ai)

A Voronoi tessellation of S2

Tessellation properties
For each e>0, There is a Voronoi tessellation such that Each cell contains a disk of radius e and is contained in a disk of radius 2e We will use a Voronoi tessellation for which : Every Voronoi cell contains a disk of area 100logn/n . Let r(n) be its radius Every Voronoi cell is contained in a disk of radius 2r(n) For a1, choose any point in S^2. Suppose there are already a1…ap that each two are distant at least 2e. Two cases: If there is a point x such that D(x,e) not intersecting any of D(ai,e) add x Otherwise stop It finished because it removes area of disk with radii e>0 When stop, we have a set that each are at list 2e apart, and all other points on s^2 are within a distance of 2e from one of them. Note that the area of a disk of radius r on S^2 is less than pi*r^2

Adjacent cells are two cells that share a common point. We will choose the range of transmission r(n) so that: With this range, every node in a cell is within a distance r(n) from every node in its own cell or adjacent cell 8r(n) 2r(n)

Theorem 4.1 For Random Networks on in the Protocol Model, there is a deterministic constant c > 0 such that bits per second is feasible whp For Physical Model, there are c’, c” such that is feasible whp

Proof: Lemma: in the Protocol Model there is a schedule for transmitting packets such that in every (1+ ) slots, each cell in the tessellation gets one slot in which to transmit, and such that all transmissions are successfully received within a distance r(n) from their transmitters From the above lemma, the rate at which each cell gets to transmit is W/(1+ ) per second

Lemma: There is a δ’(n)→0 such that Prob ( (Traffic needing to be carried by cell V) ≤ c5λ(n) ) ≥ 1- δ’(n) From the above lemma, the rate at which each cell needs to transmit is less than c5λ(n) whp. With high probability, this rate can be accommodated by all cells if it is less than the available rate, i.e., if

Within a cell, the traffic to be handled by the entire cell can be handled by any one node in the cell, since each node can transmit at rate W bits per second whenever necessary Lemma: Every cell in has no more than interfering neighbors. depends only on ∆ and grows no faster than linearly in (1+∆)^2 Thus, for Protocol Model, is feasible whp

Lemma: if ∆ is chosen to satisfy
then the above result of Protocol Model also holds for Physical Model Plug the expression of ∆ into , we get

Theorem 5.1 For Random Networks on under the Protocol Model, there is a deterministic c’ < ∞ such that

Proof: Lemma: the number of simultaneous transmission on any particular channel is no more than in the Protocol Model Let L denote the mean length of the path of packets, then the mean number of hops taken by a packet is at least

Since each source generates λ(n) bits per second, there are n sources and each bit needs to be relayed on average by at least nodes, so the total number of bits per second needs to be at least Also, each transmission over a single channel is of W bits per second, so from the above lemma the number of bits can not be more than bits per second, so

Lemma: the asymptotic probability that graph G(n, r(n)) has an isolated node and is disconnected is strictly positive if and By the definition of feasible throughput, the absence of isolated node is a necessary condition for feasibility of any throughput. Thus, is necessary to guarantee connectivity whp. We obtain the upper bound for Protocol Model

Conclusion Implication for design Not considered Number of nodes
Signal decay rate Not considered Delay Mobility

Thanks ~ 

A graph of degree no more than c1 can have its vertices colored by using no more than (1+c1) colors
So color the graph such that no two interfering neighbors have the same color, so in each slot all the nodes with the same color transmit

If V’ is an interfering neighbor of V, then V’ and similarly every other interfering neighbor, must be contained within a common large disk D of radius 6r(n)+ (2+D)r(n)