Senior Design Project OPNET Modeler ShortRange Wireless Routing

Senior Design Project OPNET Modeler & Short-Range Wireless Routing Erin Butler & Emmy Lai Advisor: Professor H. C. Chang

Grant Funded by a grant from CRA-W, a subset of the National Science Foundation,

Objective of Project Research Existing Wireless Routing Algorithms Design and Implement these Algorithms Explore OPNET Modeler l A network technology development environment Simulate a Variety of Networks, Observing Performance Metrics

OPNET Modeler Three Main Domains Network l Node l Process l

Network Domain Subnetworks l Encapsulates other network objects Communication Nodes l Model network objects with definable internal structure Communication Links l Mechanism to transport information between nodes Fixed, Mobile, Satellite Variations

Node Domain Node Model defines the internal structure of the communication nodes Node Modules l l l Processor: primary building block, sends/receives packets, overall processing Queue: extended functionality of processor, array of internal resources, subqueues Transmitter: interface between internal packet streams & external communication links Receiver: interface between external communication links & internal packet streams Connections: l l l Packet Stream: support flow of data between modules Statistic Wires: support transmission of numerical state information Logical Associations: bind two modules, allowing them to perform function together

Process Domain The Process Model defines the behavior of the processor and queue modules Interrupt Driven Execution: l l Caused by the invocation of an event Alternating Blocked and Active states Dynamic Processes: l l l Processes invoked by other processes Share Memory Architecture Parent-Child establish pair establish block of memory for two-way communication

Process Domain (cont. ) Dynamic Library operations State Transition Diagrams: l State: a mode the process can enter, state information l l Enter & Exit Executives Unforced state: wait for interrupt Forced state: continual execution of state Transition: possible movements of a process from state to state l Source & destination state, condition & executive expression Input & Output Streams

Data Analysis Tool Graphs l Statistics l Output Scalar Files: data collected in vector files during a simulation run, combine results from multiple simulations l

OPNET Editors Project Editor Node Editor Process Model Editor Link Model Path Editor Packet Format Editor Antenna Pattern Editor Interface Control Information Editor Probability Density Function Editor Probe Editor Simulation Tool Analysis Tool Filter Editor

Project Editor Main staging area for creating a network simulation Create a network model using models from the standard library Collect statistics about the network Run the simulation View Results

Node Editor Define the behavior, which is defined by modules, of each network object A network object is made up of multiple modules defining its behavior Each module models some internal aspect of the node behavior (ex: data creation/storage)

Process Model Editor Create process models which control the underlying functionality of the node models created in node editor Represented by finite state machines Created with icons that represent states and lines that represent transitions between states Operations performed in each state or for a transition are described in embedded c or c++ code blocks

Link Model Create new types of link objects Each new type of link can have different attribute interfaces and representation

Path Editor Create new path objects which define a traffic route Any protocol model that uses logical connections or virtual circuits (MPLS, ATM, Frame Relay…etc) can use paths to route traffic

Packet Format Editor Defines the internal structure of a packet as a set of fields A packet format contains one or more fields, represented in the editor as colored rectangular boxes Size of the box is proportional to the number of bits specified as the field’s size

Antenna Pattern Editor Models the direction dependent gain properties of antennas Gain patterns are used to determine gain values, given knowledge of the relative positions of nodes

Interface Control Information Editor Defines the internal structure of ICIs (Interface Control Information) which are used to formalize interruptbased inter-process communication

Probability Density Function Editor Describes the spread of probability over a range of possible outcomes Models the likelihoods associated with packet interarrival times Models the probability of transmission errors

Probe Editor Specifies the statistic to be collected during simulation Sets additional characteristics of each probe Different probes collect different statistics including global statistics, link statistics, node statistics, attribute statistics, and several types of animation statistics.

Simulation Tool Specifies additional simulation constraints Simulation sequences are represented by simulation icons which contain a set of attributes that control that simulation’s runtime characteristics

Analysis Tool Creates scalar graphs and parametric studies Defines templates to which statistical data is applied Creates analysis configurations

Filter Editor Creates additional filters on top of the ones that are already provided by OPNET Built by combining existing models with each other

OPNET Overview Layout

Carrier Sense Multiple Access Protocol (CSMA) Protocols in which stations listens for a carrier or transmission and act accordingly Three versions of CSMA 1 persistent l non persistent l p persistent l

1 – persistent CSMA When a station has data to send, it first listens to the channel to see if anyone else is transmitting at that moment If the channel is busy, the station waits until it becomes idle l When the station detects an idle channel, it transmits a frame If a collision occurs, the station waits a random amount of time and starts all over again Transmits with a probability of 1 whenever it finds the channel idle

Non-persistent CSMA Attempts to be less greedy than 1 -persistent Before sending, a station senses the channel l l If no one else is sending, the station begins doing so itself If the channel is already in use, it waits a random period of time and then repeats the algorithm Leads to a better channel utilization and longer delays than 1 -persistent

p-persistent CSMA Applies to slotted channels When a station becomes ready to send, it senses the channel If it is idle, it transmits with a probability p A probability of q=1 -p is deferred until the next slot l l If that slot is also idle, it either transmits or defers again, with the probabilities p and q Process repeats until either the frame has been transmitted or another station has begun transmitting If another station has begun transmitting, it acts as if there had been a collision (ie, it waits a random time and starts again) If the station initially senses the channel busy, it waits until the next slot and applies the above algorithm

OPNET & CSMA Basic Components CSMA project network l Transmitter Node Model – sends packets l Receiver Node Model – performs network monitoring l

CSMA Process Model Verifies the channel is free before transmitting If channel is not free, enters the wt_free state until a “channel goes free” interrupt is received

CSMA At the node level, the statistic wire is triggered when the busy statistic changes to 0. 0. The trigger is activated by enabling the wire’s falling edge trigger attribute

CSMA Scenario CSMA network model 20 transmitter nodes Uses transmitter nodes designed previously in node editor Network is ready for simulation

CSMA Simulation Change attributes to run simulation l l l Duration time Seed Value per statistics In this case, for the CSMA, seed is changed to 11

CSMA Simulation Results Graph of channel throughput S vs. channel traffic G Achieves maximum throughput at about 0. 5

Comparing Protocols (CSMA vs. Aloha)

Introduction to Dynamic Wireless Networks These networks consist of mobile hosts that communicate to one another over wireless links without any static network interaction l Due to the limited range of wireless transceivers, mobile hosts’ communication links only implemented in their geographic reason Need for a complex network to handle and maintain the forwarding of data packets

Previous Work & Routing Standards Set of Conventional Standards l l Simplicity Loop-free Low Convergence time Low computation & transmission overhead Problems in terms of Dynamic Networks l l l Frequent broadcast cause high overhead due to changing topology Heavy computational burden Limited bandwidth in wireless networks

Temporally Ordered Routing Algorithm (TORA) A network routing protocol which has been designed for use in Mobile Wireless Networks Envisioned as a collection of routers which are free to move about arbitrarily Routers are equipped with wireless receivers/transmitters Status of communication links between routers is a function of their positions, transmission power levels, antenna patterns, cochannel interference levels…. etc Designed to minimize reaction to topological changes

Properties that makes TORA well suited for use in the mobile wireless networking environment Executes distributedly Provides loop-free routes Provides multiple routes (to alleviate congestion) Establishes routes quickly (so as to be used before the topology changes) Minimize algorithmic reactions/communication overhead (to conserve available BW and increase adaptability)

Methods to minimize overhead & maximize routing efficiency Establish routes only when necessary by constructing a direct acyclic graph rooted at the destination using a query/reply process React to link failure only when necessary (ex: when a node loses its last downstream link) Scope of failure reactions minimized (ie: the number of nodes that must participate) No reaction to link activation

TORA’s Link Reversal Algorithm When a node has no downstream links, it reverses the direction of one or more links Links are directed based on a metric, maintained by nodes in the network, that can conceptually be viewed as a height Goals: l l l Discover routes on demand Provide multiple routes to a destination Establish routes quickly Minimize overhead Make shortest path routing of second importance

TORA Basic Functions Creating Routes – Query/Reply on demand l l Query packet (QRY) is flooded through network Update packet (UPD) propagates back if routes exist Maintaining Routes – Base on “link-reversal” algorithm l UPD packets reorient the route structure Erasing Routes – l Clear packet (CLR) is flooded through network to erase invalid routes

TORA FSM

Create Route State Diagram

Maintain Route State Diagram

Erase Route State Diagram

TORA Application A separate copy of TORA is run at each node Node adjust height at a discovery of an invalid route Node without neighbor of finite height with respect to destination, attempts to find new route Sends CLR packet upon network partition Exchange of UPD packets

TORA Application cont… Complete path can be found using distance table Each router maintains its own information with respect to its neighbor

Node Height Each height table for a node contains the following information l Hi = ( i, oid, ri, i, i) l I = time tag l oid = originator ID l ri = bit used to divide each reference level into 2 sublevels l I = integer used to order nodes l I = unique identifier of node Height of each node (except for the destination) is initially set to NULL: Hi = ( -, -, i)

Route Creation (-, -, A) (-, -, B) (-, -, C) (0, 0, 0, 1, E) (0, 0, 0, 2, D) (0, 0, 0, 2, G) (0, 0, F) (0, 0, 0, 1, H) DEST

Route Creation (0, 0, 0, 3, A) (0, 0, 0, 2, B) (0, 0, 0, 3, C) (0, 0, 0, 1, E) (0, 0, 0, 2, D) (0, 0, 0, 2, G) (0, 0, F) (0, 0, 0, 1, H) DEST

Route Creation Complete (0, 0, 0, 3, A) (0, 0, 0, 2, B) (0, 0, 0, 3, C) (0, 0, 0, 1, E) (0, 0, 0, 2, D) (0, 0, 0, 2, G) (0, 0, F) (0, 0, 0, 1, H) DEST

Cluster-Based Algorithm for Dynamic Network Routing Main objective: To replace individual nodes (mobile hosts) with a cluster l Lower overhead during topology changes Basic algorithm l l Divide the graph into a number of overlapping clusters Change in topology = change in cluster membership

Cluster Definitions Graphs: an organization of nodes or mobile hosts Node: list of neighbors, list of clusters it belongs to, list of boundary nodes Boundary nodes: connection from one cluster to another Clusters: l Size of cluster C, S(C) = number of nodes l Edges: edges between the nodes that are members of the clusters l Cluster-connected graph: union of clusters covers the whole graph, a path from exists from each node to every other in the graph l Redundant cluster, if removed, does not affect the connection between a pair

Four Main Topology Changes H 5 turns ON H 6 turns OFF HA connects to HB HA disconnects from HB

Procedures & Data Structures Two main Procedures: Switch ON, OFF l Each execute a similar algorithm of cluster manipulation l l l Create. Cluster Find. Essential Find. Redundant Data Structures: l l Clus_List: list of clusters Bound_List: list of boundary nodes Switch

Performance Metrics Time Complexity l Number of steps for a network to form after topology change Communication Complexity l Number of messages required to form new network Routing Overhead l Ratio of path length between source and destination

Dynamic Cluster Implementation Bottom-Up approach Process Model – defines behavior of nodes, state diagram and transitions l Node Model – contains objects (receiver, transmitter, processor) consisting of process models l Network Model – overall topology l

Process Models Decision “Tree”: Switch ON or Switch OFF

Switch ON State Diagram

Switch OFF State Diagram

Node Model Wireless LAN Model: MAC sublayer, Physical Layers

Network Model

C Language Implementation: Main Functions & Data Structures Three Basic Functions Create. Cluster l Find. Essential l Find. Redundant l Globally Declared Data Structures Neighbor. List: list of all neighbors of each node l Clus_List: contains list of clusters l

C Language Implementation: Create. Cluster Input Parameter: node ID Using the node ID and its Neighbor. List, determines all existing clusters

C Language Implementation: Find. Essential Input Parameter: node ID Ensures that no one node, except the node ID, exists in more than one “essential” class Design: series of reiterative looping and comparisons l If all nodes are found in another “essential” cluster, cluster is marked “non-essential”

C Language Implementation: Find. Redundant Input Parameter: node ID Uses “essential” list Determines whether or not the removal of a cluster affects the clusterconnectivity between any pair of nodes Output is the final Clus_List for the node ID

C Language Implementation: Simulation (initial)

C Language Implementation: Simulation (H 5 ON, node. ID=0)

C Language Implementation: Simulation (H 2 OFF, node 5)

C Language Implementation: Simulation (H 3 disconnect H 4, node 0)

AND NOW…GRADUATION!!