Document Details

FastGrowingJudgment

Uploaded by FastGrowingJudgment

Sharjah Women's College

Tags

wireless networks wlan ieee 802.11 networking

Summary

This document provides notes on wireless local area networks (WLANs), covering various aspects such as characteristics, architecture, physical layer standards (IEEE 802.11), and challenges. The document also explores different aspects of WLAN technologies. The summary is useful for students studying networking concepts.

Full Transcript

Chapter 6: Wireless Local Area Networks (WLANS) • IEEE 802.11 covers both the Physical and & Data Link Layers • Wi-Fi stands for Wireless Fidelity WLAN characteristics: • Attenuation: The strength of the signals weakens as it disperses in all directions • Interference: A receiver may receive signals...

Chapter 6: Wireless Local Area Networks (WLANS) • IEEE 802.11 covers both the Physical and & Data Link Layers • Wi-Fi stands for Wireless Fidelity WLAN characteristics: • Attenuation: The strength of the signals weakens as it disperses in all directions • Interference: A receiver may receive signals from a non-intended sender that is using the same frequency band • Multipath propagation: A receiver receiving more than one signal from the same sender • Error: Errors are more serious in a wireless network than a wires network IEEE 802.11 Architecture: • Basic Service Set (BSS): Stationary or mobile wireless station with an optional Access Point • Extended Service Set (ESS): Two or more BSS with Access points Ad Hoc Mode: • • • • P2P mode: Nodes talk to each other directly Limited backhaul Network setup on the fly Examples: UAV, Disaster response communication, military Infrastructure mode: • • • • Nodes talk to an AP AP provides backhaul Needs infrastructure element to enable connectivity Examples: Home/Office networks, airports, mall IEEE 802.11 Physical Layer I. II. Legacy 802.11: • Diffused infrared transmission • FHSS and DSSS • Operates in 2.5 GHz 802.11a: Introduced OFDM • Uses different modulation techniques • Big jump in data rate III. 11g – (Popularity of 11b & speed of 11a) • Operates in 2.5GHz, backwards compatible with 11b IV. 802.11n: Exploited larger channels (20 or 40MHz and MIMO) • 2.4 & 5 GHz, backwards compatible with 802.11 • 802.11 MIMO: Devices capable of Beamforming (directed transmission) • 802.11 MIMO: Spatial multiplexing: frames broken up and sent as multiple parts from different radios V. 802.11ac: Introduced channel bonding & MU-MIMO: • Uses Beamforming & MU-MIMO in Downlink • Supports up to 4 users simultaneously • Supports 256 QAM VI. 802.11ax Beamforming & MU-MIMO in Uplink Supports OFDMA in Uplink Supports up to 8 users simultaneously Supports enhanced Spatial Reuse (SR) & 1024 QAM • • • • Analogy: • 802.11: Long line of customers and 1 bank teller • MU-MIMO: 4 tellers serving 4 lines of customers • OFDMA: Each teller server 4 customer simultaneously Network Allocation Vector: A timer that shows how much time must pass before stations are allowed to check channel for idleness MAC Sublayer: I. II. Distributed Coordination Function (DCF): • Devices listen then talk • Receiver acknowledges upon receipt • Provides option for reservation Point Coordination Function (PCF): optional access method implemented in an infrastructure network (Not for Ad Hoc mode) • Used for time sensitive transmission • Centralized & contention free access • Stations are polled one after another • Since PCF is prioritized over DCF, stations that use only DCF may not gain access to the medium. • Super Frame is a repetition interval protocol that counters the aforementioned issue by covering contention-free PCF and contention-based DCF traffic • A special control frame called a beacon frame is initiated at the start of each repetition interval • When the station hears the beacon frame, the NAV starts the contention-free period Association / Reassociation: I. II. Passive Scan: 1. Client listens on each channel for a finite period 2. Listens for a Beacon periodically (100ms) 3. Frame includes AP’s SSID & BSSID 4. Client follows authentication procedure and joins network Active Scan: 1. Client sends probe request on each channel 2. Waits (10-20ms) for a probe response frame from AP Reassociation happens when roaming (Make-before-break). Client reassociates with new AP which sends disassociation frame from old AP. Issues Affection WLAN Performance: I. II. III. Channel Selection/Allocation: Limited number of channels implies potential conflict Channel Allocation Considerations: • Planned setup • Traffic awareness • Coordination • Selection Channel Allocation – Wish list • Communicate across Administrative domains • Regular monitoring of Network Dynamics • Reduce channels switching & Service disruption • Simple to implement IV. Rate/Link Adaptation • IEEE provides multi-rate capability • Lower rates are robust but provides less throughput • Higher rates’ robustness depends on channel quality but provide high throughput WLAN Challenges: • • • • Wi-Fi deployments becoming increasingly dense Increase in both STAs & APs Enhancing Spatial Reuse is Paramount Introduction of multi-user operation increases throughput Dynamic Sensitivity Control (DSC): Motivated by identifying the best Clear Channel Assessment Threshold (CCATH) • Uses different metrics, Local vs Global, independent vs Coordination • Beamforming, transmit power adaptation techniques • Driven by simplicity & practicality Work in IEEE 802.11ax DSC – to be deaf to OBSS transmissions TPC – To reduce Tx power for less interference • Will hurt performance unless everyone uses TPC Uses both TPC & DSC • Increase CCA threshold to overcome OBSS Tx • Reduce Tx power when engaging in Spatial Reuse Tx to avoid interfering with OBSS Tx Chapter 7: Wireless Wide Area Networks (WWANs) – Mobile Communications Spectrum Re-use • • • • • Early Schemes (1G) used a single transmitter to cover a wide area Limited Number of Channels Large waiting lists Needs 25kHz for sufficient audio quality & guard band Can support only 40 users using 1Mhz band Frequency Re-use: Illustrated as hexagonal cells where the cell-edge boundary varies in Radio Propagation, terrain, end-point receiver sensitivity • Signal strength gradually reduces towards the edge • Edges can overlap and coverage hole at others • Adjacent cells allocate different frequencies to overcome this problem Cell size & Capacity • • • • Shrink the size to enable more re-use Small low-powered stations to cater to areas with more users Need to deploy more infrastructure Types of cells: a. Macro Cells (~10km radius, sparse areas) b. Micro Cells (~1km radius, dense areas) c. Pico Cells (areas of buildings / tunnels etc.) d. Small Cells (inside the home) Multiple Access: Allows Multiple users to access the System by utilizing s resource sharing principles • • • • • Frequency Division Multiple Access (FDMA) Time Division Multiple Access (TDMA) Code Division Multiple Access (CDMA) Orthogonal Frequency Division Multiple Access (OFDMA) Non-orthogonal Multiple Access (NOMA) Duplex Operation I. II. Simplex: Push to talk System • Cannot transmit & receive at the same time • Noticeable delay between talking and listening Duplex: Talk in both directions at the same time • Uplink (UL) & Downlink (DL) • Using Frequency Division Duplex (FDD): Tx on one frequency (f1) & and Rx on another frequency (f2) • Using Time Division Duplex (TDD): Tx and Rx on the same frequency but different timeslots. Short bursts so no noticeable delay to the user Service Continuity I. II. III. Mobile moves from one cell to another Handover to maintain service continuity • Smooth reroute without interruption • Earlier calls dropped during transition Typical Handover • User assisted (mobile best placed to monitor signals from different stations) • Network initiated (only networks know channel availability & when to initiate handover) 1st Generation Systems: 1G • Big bulky expensive phones with no standardization nor roaming. Provided limited capacity & coverage • Frequency modulation in 450/900 MHz • Analog system: Major achievement of its time • Grew rapidly leading to capacity crunch • No security features 2nd Generation Systems: 2G Frequency Bands • Original aim was to use 900 MHz band • Most successful system with 1+ billion subscribers in 2004 • Global standard: enabled international roaming & increased widespread adoption 2nd Generation Systems System Architecture: • BTS: transmits & receives signals from the MS • BSC: manages one or more BTS. Handles channel set-up security, authentication, paging & handovers • MSC: interfaces with other MSCs o Interfaces with Authentication Centre (AuC) to authenticate users o Interfaces with HLR & VLR registers to resolve location information for call routing o Coordinates handovers o Gateway MSC (GMSC) interfaces with Public Switched Telephone Network (PSTN) Equipment Identity: • • • • International Mobile Equipment Identity (IMEI) 15-digit number used to identify equipment Hardcoded into mobile during manufacture time Checked with Equipment Identity Register (EIR) when trying to access a system o Whitelisted: Access Allowed o Blacklisted: Access Blocked (stolen/unapproved device) Subscriber Identity: • 15-digit number contained in the SIM card (Subscriber Identity Card) • Enables operator to link the phone number with subscriber • Authentication Key: Stored in the SIM card, used to generate a cipher key Signaling Methods: • Control Channels: Used to page/call a mobile o Take access request o Control signals transmitted using FSK • Forward control channel acts as a beacon o For mobiles joining the network o For mobiles looking for alternatives if handoff required o Channels with highest signal strength chosen • Reverse control channel o Carries registration message from the mobile o Call setup messages when a number is dialed o Responding to paging requests from the BS during an incoming call Initialization 1. 2. 3. 4. 5. 6. MS (Mobile subscriber) scans all available frequencies Records signals from BTSs and broadcasts information Identifies strongest beacon frequency BTS forwards this to MSC via BSC MSC queries the EIR & AuC and based on the reply, authenticate the MS MS is now ready to make or receives calls Call Initiation 1. MS sends an access request over the RACH to BTS 2. If not ACK from BTS, random wait and entry 3. If BTS receiver the access message o It directs the MS to a specific radio channel where call setup can be progressed o Subsequently the call can commence Call reception 1. Caller Places the call 2. Network checks the location register for last known location of callee 3. Sends a page message on the paging channel of a group of BTSs in the region 4. When the cellee receives the paging message, it responds on the reverse control channel 5. BS allocates a radio channel and timeslot to the callee so the call can take place 6. Call continues until either party hangs up Handovers (HO) • Handovers occur when: o Drop in signal quality o Load balancing: move users from heavily loaded to lightly loaded BS • MS scans other channels looking for stronger beacons & reports back to BTS • Network Knows o Link quality between MS & BTS o Strength from other BTS at the MS o Availability of channels from nearby BTS • Network initiates handover 1. Notifies new BTS to reserve channels 2. Notifies MS to establish link with new BTS 3. New BTS confirms HO completion 4. Old BTS clears state and frees up the channels Global System for Mobile Communication Summary: • • • • • Multi-stakeholder, cross-border join initiative with billions of subscribers Transition from 1G Analog to a 2G Digital System Brings the best of both FDMA + TDMA to improve efficiency Delivered on capacity, security & global roaming Is circuit switched and has no support for packet switching 2.5G – General Packet Radio Service (GPRS) • Introduced Packet switching which improved utilization through multiplexing o Supports IP based connectivity o On-demand data • IP based operation eased migration to newer generations GPRS Architecture • • • • Builds on top of GSM architecture Introduced 2 new elements to provide data services Network between BTS and BSC unchanged Serving GPRS Support Node (SGSN) introduced to provide data services within the network o Services include authentication, location tracking, QoS monitoring. Link between MS and the GGSN • Gateway GPRS Support Node (GGSN) introduced as default gateway to the outside world o MS must attack to SSGN to receive data services o It is assigned an IP address by the GGSN Operation 1. 2. 3. 4. 5. 6. MS registers with the network at startup Transmits a burst on Random Access Channel (RACH) Identifies itself, indicates it wants to provide location update Network authenticates the MS, SGSN notes its location MS enters standby mode Monitors Packet Paging Channel (PPCH) for any incoming alerts 7. MS enters ready mode to send/receive data 8. MS sends a Packet Tx request using Packet Random Access Channel (PRACH) in the UL 9. BS sends a grant on Packet Access Grant Channel (PAGCH) 3rd Generation Systems: 3G Motivation for 3G: • • • • 2G introduced voice and text messaging 2.5G (GPRS) introduced packet switching and higher data rates Advances in other technology areas Competition: Wi-Fi Evolution to 3G: Internationalization of cellular standards Ways to support higher data rates • • • • Provide increased amount of spectrum Increase the transmit power of the transmission Decrease the distance between transmitter and receiver Increase number of antennas and utilize beamforming W-CDMA Technology • • • • Utilizes 5MHz Bandwidth unlike 1.5MHz in CDMA-2000 Better security and spectral efficiency Anti-jamming properties Neighboring cells can use the same frequency 3G UMTS Architecture • Core Network: Switching/Routing and provide transit to in/out traffic • UMTS Radio Access Network (UTRAN): Radio Network Controller (RNC) handles power control, channel allocation, RRM, handover and NODE B serves as UEs • UE Side: Mobile Equipment and a Universal SIM (USIM): supports access priority and priority users Power Control 1. BSs receives all UEs with similar power. Stronger signals from near UEs and vice versa 2. Once connected, BS measures received power for each UE during each time slots by sending a bit to the UE indicating the power to be stepped up/down 3. Change in propagation implies constant change may be inevitable Discontinuous Reception Support • • • • • Conserve battery life by putting radio to sleep Divides paging channel to groups/sub-channel Network tells UE which paging sub-channel to use UE only needs to listen to its paging channel UE turns receiver ON only when it needs to monitor the paging channel 3G-3G and 3G-2G Handover • UE moves out of range of a 3G cell and must be handed over to another frequency (Hard Handover) • UE moves from one 3G cell to another 3G cell o Adjacent cells can be on the same frequency o UE may hear from other BS simultaneously (Soft handover) • RNC makes the handover decision • Inter-System Handover o Handover from 3G cell to GSM cell 4th Generation Systems: 4G Technology Innovations in LTE • IP Core, Small Cells, OFDMA, Carrier Aggregation • LTE Unlicensed, MIMO, LTE-M Evolved Packet Core (EPC) • Base stations called evolved NodeB (eNB) • eNB connected to the core via an IP Network • eNB can talk directly to each other directly over X2 interface (not possible in previous generations) • Enables deploying eNBs anywhere there is an IP connection (ex: small cells in home (HeNBs) using the home broadband Resource Management • Provides variable number of resources in the Downlink o Not all blocks are the same at the same time for all users o Some blocks may be better at any given time • Single carrier FDMA in the Uplink o Unlike OFDMA, user allocated a contiguous part of a channel Inter Cell Interference Coordination (ICIC) • A and B high power on RB f3 will cause interference to each other • A & B exchange ICIC information directly. Avoid using RB for cell edge users LTE FDD Frame Structure • • • • • • Frame duration 10ms Each slot is 0.5ms long 20 slots in a frame, 2 slots in a sub-frame 1 Resource block (RB) = 12 sub-carriers per slot 1 sub-carrier has 15kHz bandwidth 1 RB = 12 * 15 = 180kHz bandwidth per slot Small Cells: AKA HOME Base Stations (HBS) or Home eNodeB (HeNB) • Customer Premises Equipment (CPE) • Help to Plug indoor coverage holes • Improve capacity due to extensive frequency reuse but limited coverage Carrier Aggregation: • Allocates more than 20MHz to support higher data rates o Can combine within or across bands o Like channel bonding in Wi-Fi 11n, 11ac and 11ax Multiple Input Multiple Output (MIMO): • Improves performance • Spatial Diversity: Multiple replicas received over multi-paths • Spatial Multiplexing: Transmission of multiple data streams over parallel spatial channels LTE Unlicensed (LTE-U): AKA License Assisted Access (LAA) • Access “extra” bandwidth in the DL using the unlicensed band LTE-M (for IoT): Lightweight LTE for IoT applications • Either push for higher data rate for bandwidth hungry apps or push for lower complexity/cost for low data rate apps • 1.4Mhz, low complexity, low data rate, long battery life 5th Generation Systems: 5G 5G System concept services and requirements: • Extreme Mobil Broadband (xMBB): High data rates, low latency, extreme coverage • Massive Machine Type Communication (mMTC): Scalable connectivity, Tx of small payloads, wide area coverage • Ultra-reliable Machine Type Communication (uMTC): Low latency, high availability & reliability communications Deployment examples • Massive MIMO & Ultra Dense Network (UDN) deployment with small cell areas, narrow beams, and operation at higher frequencies • Centralized macro-cells, with distributed UDN nodes mmWave Communications • Uses LTE in lower bands for coverage • 5G new air interface in higher bands • Improve network utilization and user experience • Dual connectivity o Lower bands for control signaling o Fast fallback when mmWave connectivity is lost Phantom cell concept: small cells that are controlled by macro-cells in split between C-plane and U-plane Radio Resource Management: • • • • Early generation employed TDMA, FDMA. Simple but not optimal 3G used CDMA. Provides frequency diversity LTE introduced OFDMA. Provides both time and frequency diversity Non-Orthogonal Multiple Access (NOMA) in 5G. 2-3 more connected devices & up to 50% user and system throughput gains Non-Orthogonal Multiple Access (NOMA): • Uses power domain for signal separation o Isolates signals using same resources based on power level o Different users have different path losses o Uses successive interference cancellation at the receiver • Power sharing reduces power allocated to each user o Improves system capacity and fairness • Supports more connections than legacy tech Radio Access for MTC: • Connection reservation approach used in LTE o Users transmit a preamble, BS responds for user identification o User requests connection, BS responds for subsequent transmissions Frequency Bands for 5G: • Focus on both cmWave and mmWave spectrum o 3Ghz- 30Ghz o 30Ghz – 100GHz • No terrestrial mobile services over 6Ghz • 7Ghz unlicensed spectrum in the 60Ghz band Spectrum challenges: • Too many bands • Implications on cost, for factor, Energy efficiency • Challenges due to roaming support 5G Spectrum Usage Scenarios: • Need sufficient low frequency band spectrum to satisfy coverage • Need large amounts of Spectrum in cmWave and mmWaves to satisfy high data rates • Licensed most appropriate for uMTC applications • Exclusive access, licensed shared access, unlicensed access opportunities exist Chapter 7: Wireless Wide Area Networks (WWANs) – Satellite communication Satellite Broadband Wireless • Use of satellites for personal communication is recent • Satellite use falls under three categories o Research: Acquire scientific data o View earth o Reflect signals: Satellites are used as reflectors to relay signals Satellite Transmissions: • Satellites on one of four frequency bounds o L Band: 1.53-2.7 GHz o C band: 3.6-7 GHz o Ku band: 11.7-12.7 Ghz for downlink; 14-17.8 for uplink o Ka band: 17.3-31 Ghz • Frequency Band affects the size of the antenna o TV Satellites use typically Ku band which requires a dish antenna of about 1m in diameter Satellite Classification: Satellites are classified according to the type of orbit they use • Low earth orbit (LEO) • Medium earth orbit (MEO) • High earth Orbit (HEO) Low Earth Orbit (LEO) satellites: • • • • • Circles the earth at an altitude between 200 to 900 miles Travels at high speeds Has low latency Uses low-powered terrestrial devices (RF transmitters) Round trip time: 20 to 40 milliseconds for a signal to bounce from earth to LEO and back to earth Medium Earth Orbit (MEO) satellite: • Circles the earth at an altitude between 1,500 to 1,000 miles • Advantages o Does not have to travel fast; MEO takes 12 hours to circle earth o Has bigger earth footprint • Disadvantages o The higher the orbit the higher the latency o Round trip time: 50 to 150 milliseconds High Earth Orbit (MEO) satellite or Geosynchronous Earth Obit (GEO • • • • • Stationed at 22,282 miles Orbit matches the rotation of the earth Three GEO satellites are needed to cover earth Highest latency at about 250 milliseconds Requires high-powered terrestrial transmitting devices Chapter 9: Network & Transport Layer Issues in Wireless Networks • At first nodes were stationary always on and were connected by wired links. • Unlike Wireless, routing is stable since addresses don’t change • Transport protocols adapt to changing link conditions • In wireless environment o Nodes are not necessarily stationary o Nodes may not always have access to mains power o Link conditions change due to congestion and characteristics of the wireless medium TCP/IP over Mobile/Wireless Networks • TCP/IP is versatile (used with wired or wireless) • Communicating nodes need mobility support when moving. Updates to address and routing for communication • Link characteristics may change significantly. o Adapting to transport parameters to ensure a seamless experience § In high bandwidth link: Ramp up sending rates § In low Bandwidth: Slow down sending rates § React to losses accordingly as not all of them are due to congestion IP Addressing 1. Nodes join a network which assigns IP address using Dynamic Host Configuration Protocol (DHCP). IP address consists of Network ID and Host ID 2. Applications use IP address to establish connections 3. Nodes move to another network a. Cannot use same address in new network b. Must obtain new network ID and Host ID (New IP Address) c. Will break existing connections d. If restarting connections is acceptable, that’s fine 4. Mobile IP is used to maintain existing connections but has overheads Mobile IP Analogy 1. I register with my local post office (Home address (HA)) 2. I travel to a foreign land, register at the local post office there (Foreign Address (FA)) 3. My home address is informed to forward my letter to my foreign address Mobile IP Operation 1. 2. 3. 4. Foreign Agent (FA) advertises that it can offer mobile IP service Mobile Host (MH) requests registration with FR FA relays to Home Address to seek approval Correspondent Host (CH) communicates with MH on his Home Address (CH doesn’t know MH has moved somewhere else) 5. HA tunnels communication to the MH via the Foreign Agent (FA) 6. MH directly responds to CH Mobile IP Tunneling • HA intercepts packets from CH and tunnels them to Foreign Agent • Susceptible to man-in-the-middle attack if not authenticated Routing in Multihop Networks • • • • • Problem: Find a route from the source to destination Need to identify your neighbors Each node computes shortest path Overhead involved for info sharing Adaptation of routes o Topology changes due to mobility & link changes • Need to know route, next hop, and routing metrics of next destination Destination Sequenced Distance Vector (DSDV): • How it works: Initially nodes are only aware of their neighbors • All nodes have “distance cost” for nodes • When the link is not reachable the cost will be infinity Dynamic Source Routing (DSR) • Source Routing: source discovers the route to complete path in each packet where intermediate routers only need to follow the path • DSR has two steps being o Route discovery o Route maintenance • Route discovery is when a source has to send a packet to a destination by broadcasting a route request (RREQ) to all its neighbors. All RREQ have a unique ID • Every neighbor appropriates itself to the path, increments the metric and forwards the request. • What happens in DSR: Image there is a source and a destination, and you want to send something, but you don’t know where the destination is. • Now, how do you find where the destination is? You just going to shout asking where B for example is, somebody going to hear you and going to relay this message on until the message goes to D and then D will inform you, I heard you through here and here … and therefor you know all the sequency of the nodes through which you get to D. • A node does not follow a request that it has seen • When the requests reach the destination, it picks the best path and sends the path back to the source back in a route reply (RREP) • Route maintenance: Should a node notice a large loss using the route, it will inform the source to use a new route with the next sequence no. • Advantages of DSR o Intermediate routers keep no state o Route discovery is initiated when there is a need • Disadvantages of DSR o Extra overhead in packets o First packet has large delay Adhoc On-demand Distance Vector (AODV) routing • Combines the best of DSDV and DSR • Route discovery only when needed • Distance vector type routing table entries are setup during route discovery o Every node that receiver the RREQ sets up a reverse path back to the source o Every node that receives RREP setups up routing table entries to the destination • Where RREQ and RREP serve as route advertisements in DV • Sequence number associated with source & destination to the seq no. in DSDV Which routing approach would you use in a scenario with node mobility? Why/Why not? ANS) In a dynamic environment, you cannot have a table as you’d have to frequently update it. DSR is the optimum routing solution to this issue Is flooding the network a better approach than routing? Why/Why not? ANS) It depends on the case of use. In Routing, you receive a package and forward it to the next hop. Flooding is forwarding the package to everyone, and they keep forwarding it further. Routing is better when you want to keep the spread of information to a minimum. Flooding is better when you want to get information out quickly Routing in Low power and Lossy Networks • • • • • One more root/sink node Many endpoints Convergecast traffic Root node needs to advertise presence Nodes that discover root can join its network Nodes may be battery powered Lightweight mechanisms to discover and maintain routes How does Routing in Low power and Lossy Networks occur: 1. Root creates a Destination Oriented Directed Acyclic Graph (DODAG) rooted at the sink 2. Path from a node to sink consists of edges in a DODAG 3. Each node in a DODAG is associated with a rank value 4. Rank of nodes along any path towards the sink keeps decreasing a. Nodes closer to sink have lower rank than those farther 5. Creation of a tree structure in the network 6. RPL has been widely applied, a popular application being smart metering networks 7. One or more roots send a DODAG Info. Object (DIO) - analogous to Wi-Fi Beacon! a. DIO contains root ID b. Frequency of sending DIOs decreases over time i. Less need for frequent beacons once the network stabilises 8. Each node that receives a DIO will do some processing and forward the DIO 9. Nodes may hear multiple DIOs a. DIO from same root forwarded by its neighbours who may have received it b. From other root nodes that may be sending their own DIO 10. Upon receiving a DIO, each node computes rank and stores this along with node ID of the node that sent this DIO & forwards the DIO message a. Different rank values for different DIOs that it may receive b. Chooses to join the network of the root for which its rank is lowest i. Chooses the node that sent it this DIO as its default parent c. May also maintain information on alternative parents (if any) 11. Node then sends DODAG Advertisement Object (DAO) to the root via default parent a. This sets up the route in the reverse path (Uplink) 12. A new node (N) that appears in the network needs a DIO to join a network 13. It sends a broadcast message called DODAG Information Solicitation (DIS) a. Neighbours that receive the DIS, respond with the most recent DIO message that was received from the root 14. When N receives DIOs, it computes rank for each, chooses a default parent (one for which its rank is lowest) and joins the tree 15. Tree structure may change depending on changes in the network • Transport Protocol Overview: The transport layer is responsible for processto-process delivery of a message, from one process to another. • Packet delivery using IP & port Addresses o Destination IP defines the host where the port number will define on of the processes on the host TCP Congestion Control • TCP maintains a Window Size Variable (W) o Additive increase: when no loss o Multiplicative decrease: When ACK is not received • If packet is lost slow down. Losses are due to congestion, poor channel quality or corruption from interference TCP Performance using Mobile IP • Triangular routing (CH-HA-FA) can increase round trip time (RTT) o Longer time to get an ACK o Can lead to slowing down if sending has ACK timeout • TCP throughput varies with RTT o High RTT, low throughput o Low RTT, high throughput • Impacts on mobility include: bursty losses, patchy signals In a nutshell • TCP backs off on packet loss implicitly assuming congestion o If loss due to ‘Congestion’ § Back-off makes sense, relieves congestion o If loss due to inherent characteristics of the wireless medium § Back-off unnecessary, degrades throughput • Ideally o TCP should back-off on Congestion and not otherwise • How to achieve this? o Detect loss (duplicate ACKs / Re-transmission Time Out RTO) o Classify Loss o React accordingly Solution approaches: Goal: Hide wireless losses from send • Network based solution o Split connection approach to wireless and wired leg • Endpoint based solution o Explicit Loss notification o Inter/predict type of loss at the send Other issues to consider • How to deal with mismatch in rates between two networks • TCP’s window adaptation may be suboptimal in such cases • Maintain different window sizes for different links Chapter 10: Wireless for the Internet of Things (IoT) What is IoT? • “Smart-X” • Low cost, small form factor Categories of IoT • Industrial IoT o Transportation o Aircraft engines o Smart cities o Light • Consumer IoT o Wearable o TV o Phones o Home monitoring What’s causing this revolution? (Business Perspective) • Improve operational efficiencies • Lower cost, better experience, new revenue stream • Connected products and sustained engagement with customer What’s causing this revolution? (Technical Perspective) • Advances in hardware • IP has become all pervasive (connectivity is everyone) • Low power, low range, and ubiquitous connectivity solutions Typical Endpoint Hardware Architecture: • Microcontroller (ex: TI MSP430, Atmel ATMega) o For embedded applications o Low power consumption • Memory (few KB of RAM) • Low power radio transceiver (ex: CC2420, CC1000) o Operates on ISM band o Low bandwidth, power consumption, data rate • Sensors • Batteries Typical Endpoint Software Architecture: Refer to slide 11 in Chapter 10 for the figure • From an OS perspective, the goal is to strip down functionality that creates overhead • Emergence of IoT platforms • One use case of IoT the Libelium smart world Radio Frequency Identification (RFID) • Low cost, power radio frequency system to transmit information through radio waves • Used to identifying objects • Comprises of a tag, antenna, reader, and software RFID Tags (AKA Transponders) • • • • • Comprises of a transmitted and responder Include integrates circuit, non-volatile memory, and a simple microprocessor RFID can store data Reader: device that captures and processes data received from the tags Passive Tags (most common type) o Small, cheap to produce o Does not require battery, uses RF waves • Active Tags o Has battery, can transmit signal farther o Limited lifetime o Beacons transmit on a periodic basis Readers (AKA Interrogators) • Devices that connect to a network that transfers data obtained from tags to a computer • Some readers can write data onto tags • Can provides energy that activates passive tags • Read distance is determined by the size, location & power transmitted of the tag • Different frequency bands to cater to different types of applications LPWA Design Goals • Long range • Low power consumption and cost • Scalability Long Range Operation • Range: 1 – 50 KM • Very high Receiver sensitivity of -130 to 145 dBm • Slower modulation: lower data rate Use of Sub-GHz Spectrum • We can use unlicensed frequency bands less than 1GHz to keep the cost low • Popular unlicensed spectrum in sub-GHz bands: o SIGFOX, LORA, Telensa, Weightless Modulation Techniques • Ultra-Narrow Band (UNB): o Uses a very narrow band o Simple, inexpensive antennas o Used by SIGFOX • Spread Spectrum o Uses the entire band o Resilient to interference & jamming o Used by Lora Network Topology Mesh Topology (Not popular, requires infrastructure) Short-ranged wireless Star Topology (Popular, simple) No support for “Place and Play” “Place and Play” More unfractured cost Saves power LPWA up to 50KM Duty Cycled Operation • In Uplink: battery-powered end-devices turn on transceivers • In Downlink: communications are scheduled unless end-devices are always listening • Between Uplink and Downlink, listen for a reply from base station • Class A o Most energy efficient o Downlink available only after sensor Tx • Class B o Somewhat energy efficient o Latency controlled downlink • Class C o Least energy efficient o No latency for downlink communication Medium Access Control • Lightweight Channel Access Methods o Listen to medium and wait if someone is transmitting • ALOHA o You transmit, you don’t listen • It is better to use ALOHA protocol in LPWA Dumb Devices, Intelligent Backend • Dumb devices: sends messages to all gateways in radio reach • Intelligent Backends: Suppresses duplicates, security checks, adaptive data rate, acknowledgement scheduling Low Cost Vital • • • • • Cheap (<5$) Low connectivity costs Uses unlicensed band Minimum infrastructure Strips silicon off end device Popular Players - SIGFOX • • • • • • • Frequency: 100 Hz Uses Star topology Device Cost <= 1$, Subscription: 1-15$ Business Model: end-to-end service provider Uplink: <=140 12-byte messages, Downlink: <=4 8-byte messages Data rate: 10 bps-1kbps Power consumption: 50 microwatts Popular Players – LoRa Alliance • • • • • • Frequency: Chirp Spread Spectrum Uses Star topology Device Cost <= 5$ Business Model: Private and public (through MNOs) Data rate: 50kbps (FSK) 11kbps (LoRa) LoRa: patented by Semtech Popular Players - InGenu • Began as On-Ramp Wireless targeting smart Metering/Utilities market • IEEE 802.15.4k Standard, 2.4 GHz (not sub-GHz) • Patented technology: Random Phase Multiple Access • Claims: o Range up to 60km o 20 years battery life o 1-mile underground penetration SIGFOX Technology Operation • Devices transmit using BPSK over 100Hz and sleep for most of the time o Ultra-narrow band – long range but low data rate o No more than 140 messages/day/uplink no longer than 12 bytes o Allows only 4 messages/day/downlink each no longer than 8 bytes • Frequency diversity • Franchisee based business model LoRa/LoRaWAN operation • Uses Chirp Spread Spectrum at the PHY layer and offers receiver sensitivity close to -135 dBm • Multiple transmission parameters • LoRaWAN MAC protocol o Support classes A, B, and C • Anyone can buy LoRa and build their own network Key Challenges • Scale, Interference, Interoperability between technologies • Transition path, Authentication, Security, Privacy • Mobility & Roaming support Summary - LPWA • LPWA is 1/4th of IoT connections • Competitors are LTE-Cat M, NB-IoT, 802.11ah Implications of technology choices • 1-tier solution (Sub-GHz ISM, cellular) • 2-teir solution (Short range + long range radios) Connectivity Consideration • Deployment Density à how many devices are you going to deploy? • Deployment Topology à Are you going to deploy a star topology? • Failure Detection & Recovery à How your sensors will notify if something when wrong? • Technology Options à what are options that are available to you? Typical Dilemmas • Build vs Buy • Manage oneself vs outsource • Cost Vs performance IoT- Current State of Play • Multiple vertical solutions – little/no interoperability • Sustainable IoT • Too many competing alternatives Diverse Ecosystem: • Chip/ Device / Module vendors • Connectivity, platform providers • Applications / Analytics solution providers

Use Quizgecko on...
Browser
Browser