Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 1 of 201
`Case 1:16-cv—02690-AT Document 121-13 Filed 08/05/16 Page 1 of 201
`
`E-3
`
`E-3
`
`
`Case 1:16-cv—02690-AT Document 121-13 Filed 08/05/16 Page 1 of 201
`
`E-3
`
`E-3
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 2 of 201
`
`Network Analysis Corpomtion
`
`A hop-by-hop passive echo acknowledgement
`(HBH Echo Ack) along the path. When device i trans-
`mits a packet, it waits a sufficient time to allow
`devxces that receive the packet to repeat it. When
`any of these repeats the packet and the packet is
`received by device i, it considers it as an Ack.
`
`In a Foint-to-poxnt network, such as the ARPANET, the channels
`are f.xed so that when node J recexves a packet on channel k it
`knows the device, say i, whxch has transmitted the packet to it
`and thus can transmit a specific HBH Ack to device i. m the
`Packet Radio System, however, this information is not available
`Therefore, the HBH Ack must be independent of g» p^H thjt the'
`2^e^travels_on. The HBH Echo Ack test used included"^
`following:
`
`1)
`
`identification of the packet
`
`2) tests that the MHN of the Echo received la
`
`SSiif SS S£ MHVf .the packet ^crläe\^s
`™Vt lu tre Packet has advanced along the
`path, rather than being a retransmission from
`devices that had the packet previously.
`
`The HBH Echo *nk has several advantages over a specific
`
`HBH Ack
`
`iire/L3^1^63 Si* rePeater (hardware and soft-
`Trl f° ^'^ lt need not constract and manaqe
`acknowledge.nent packets. ""»neige
`
`2) it reduces the traffic overhead of transmittina
`specxfxc acknowledgements. This is most significan?
`in a broadcast network. »-Lynxx.i cant
`
`7.21
`
`^-——--a-.^Bu—
`
`._._____-____„,
`
`
`
`F^WVHVnpnnw^M
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 3 of 201
`
`Network Analysis Corporation
`
`3) it enables acknowledgement of several devices at
`a time; in particular, all devices which store the
`packet with a MHN larger than that received are
`acknowledged.
`
`4) It enables shortening the transmission path, as
`described below.
`
`Since the RSP's by repeaters are randomized in time, a terminal
`frequently does not identify the repeater nearest to the station
`within range of the terminal. In fact, if two repeaters are labelled
`on the same path to the station and both are within range of the
`terminal, there is a higher probability that the terminal will identify
`the repeater farther away from the station since, on the average, the
`latter handles less traffic. Suppose a single data rate channel
`is used throughout the transmission path between terminal and
`station, the terminal identifies a packet transmitted to it by
`its specific terminal ID, and the station can recognize any packet
`destined for it. Then a communication path as shown in Figure 2
`may be established. In Figure 2, the terminal is within an effective
`range to R4, R5, and R6; and the station within an effective range
`to Rl, and R2. The terminal shown identified R6 as the repeater
`to which it transmits. The path from termianl to station will
`
`usually be , T-R6->R5->R^R3 *R2 *S; and from the station to the terminal
`S.R2-*R3+R4-*T. The end devices, terminal and station, transmit the
`Echo Ack iimediately after receiving the packet, and transmit it with
`MHNfO; thus they acknowledge all devices which still store the
`packet. In particular, when R4 transmits a packet towards the
`terminal it is addressed to R5, however, the terminal may receive
`this packet and acknowledge both R4 and R5 simultaneously. Similarly,
`when R2 transmits towards the station it addresses the packets to
`Rl, when the packet is received by the station It acknowledges
`both Rl and R2 simultaneously.
`
`7.22
`
`
`
`'■'
`
`•' ■um ■■■■
`
`■P1'1
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 4 of 201
`
`'rm-
`
`Nctuork Analysis Corporation
`
`D. Performance Measures
`
`Throughput
`
`Considering the set of stations and the set of terminals
`as the end devices, the system throughput (in packets) is de-
`fined as the rate of information packets (IP's) that originated
`at stations and arrived at terminals, plus the rate of IP's
`that originated at terminals and arrived at stations.
`
`Delays
`
`Thd following delays are measured:
`
`1) Terminal delay to identify specific repeater.
`
`2) Termina] delay to establish communication with
`station and to negotiate protocols.
`
`3)
`
`End-to-end delay for an IP.
`
`4) Terminal interaction delay as a function of the
`number of IP's transmitted and received. The interaction
`delay is defined as the time elapsed from terminal
`origination to departure.
`
`Blocking and Loss
`
`When a terminal aoes not successfully identify a repeater
`(or station) after transmitting an S^ for the maximum number
`of times specified, it is considered blocked. In addition,
`
`7.23
`
`—
`
`
`
`^^m^^gmmmKwmmmmm^^f^m^^
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 5 of 201
`
`Network Analysis Corporation
`
`under certain conditions, terminals wil. depart from the
`
`system without completing communication. This will contribute
`to additional system loss. The blocking and loss are measured
`separately since the former indicates the difficulty m
`entering the system, whereas the latter reflects on the
`inefficiency of the routing.
`
`Device Performance
`
`1) Probability that the station is busy. The station
`is sampled during the simulation and is assumed busy
`
`if it is actively receiving or transmitting; otherwise,
`it is assumed idle. This measure is an indication of
`the channel traffic at the station.
`
`2) Successful completions by repeaters. The number
`of packets that each ren-ater has successfully switched
`are counted. This indicates the distribution of load
`in the network and reflects on the power duty cycle
`of repeaters.
`
`Other Measures
`
`1) The number of terminals in the system, the total
`number of packets stored in the system, the number of
`events to be processed; all as a function of time.
`
`2) The complete output includes a detailed description
`of the flow of significant communication events.
`
`7.24
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 6 of 201
`
`l^l&wm^mm^mv^*
`
`r- '"•■■
`
` '- — I i w ■ ■
`
`MM
`
`Network Analysis Corporation
`
`TYPICAL COMMUNICATIONS PATH
`
`FIGURE 2: The solid lines indicate the labelled path
`between the repeaters and station. The dashed lines in-
`dicate the effective connectivity of terminal and station
`to repeaters. The arrows indicate the practical trans-
`mission patn for the particular terminal, to and from the
`station.
`
`7.25
`
`
`
`
`
`■^■»^^"« m i • "■"
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 7 of 201
`
`Network Analysis Corporation
`
`VII. LOGICAL OPERATION OF DEVICES
`
`A. States of Devices
`
`Each device is characterized by a state vector. Some of
`the state variables will be needed in the physical devices, for ex-
`ample, the label, a parameter indicating the maximum number of trans-
`luissions, the maximum handover number to be assigned by repeater and
`station, the state of occupancy of its storage, etc. Other variables
`are particular to the simulation. The following are examples:
`
`Operational State of Device
`
`PR - Passive Receive State: The device is in receive state
`and does not sense any carrier.
`
`AR - Acrive Receive State.
`
`AT - Active Transmit State.
`
`ART- Active Receive and Transmit.
`
`When a device is in state AR or ART, it can be receiving
`several overlapping packets simultaneously. In the program, we use
`a common channel configuration (half duplex). Thus, since carrier
`sense is used for channel access, the device can change to AT only
`from PR and to ART only from AT.
`
`7.26
`
`
`
`"^
`^■IPWWi^J»»"'"""11 " ■ 11 i.i i »
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 8 of 201
`
`Network Analysts Corporation
`
`Number of Overlapping Packets
`
`This number is incremented by one whenever the device is
`in state AR of ART and a new packet begins to arrive; and
`decremented when a packet transmission ends. The number of
`overlapping packets indicates the number of times an end of
`packet transmission has to occur before the device changes its
`state to PR.
`
`End of Busy Period
`
`This time is recorded for the purpose of saving CPU time.
`The transmission time of a packet is set to the End of Busy
`Period plus a random tire; otherwise the devide may be called
`to transmit a packet several times during its busy period.
`
`B. Terminal
`
`When a terminal originates a message, it begins to transmit
`and retransmit a SP to identify a specific receiver. If it does not
`identify a specific receiver after a specified number of transmissions,
`it departs from the system. We say that such a terminal is locked.
`When a terminal identifies a specific receiver, it substitutes the
`label and MHu' sent by the receiver itno its IP and begins to transmit
`its IP. The IP is retransmitted after short waiting periods of time
`until a HBH Echo Ack is received. At that time, the terminal stores
`the IP iior a longer period of waiting afr.er which the IP is reactivated
`if an ETE Ack is not received.* The terminal is expecting several IP's
`
`* We use the term retransmission when a device waits a relatively short
`period of time (less then 2 IP slots) and is awaiting a HBH acKnowledge-
`ment. We say that a packet is reactivated when an end device stores
`the packet, awaiting an ETE Ack. ;:hen a packet is reactivated, it goes
`through the whole process of retransmissions.
`
`7.27
`
`^__^_11__
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 9 of 201
`■■■^•■■■■■■■■■■■■■■■■■■ni
`
`Network Analysis Corporation
`
`from the station, which are ETE acknowledged by the terminal. When
`all IP's from the station are received and ETE acknowledged, the
`terminal departs from the system.
`
`C. Repeater
`
`A repeater does not distinguish between IP's or ETE Acks,
`except for their transmisison time. When an IP (or ETE Ack) is re-
`crived by a repeater (addressed to it) and the repeater has available
`storage, it stores the packet, decrements the MHN, modifies the packet
`label according to the routing, and begins to transmit and retransmit
`the packet, awaiting the Echo Ack. When an Echo Ack is received, the
`repeater discards the packet. When a repeater is not successful in
`transmitting along the "shortest" path, it begins to search for an
`alternate receiver by transmitting SP's. When one is found, it trans-
`mits the entire packet to it; otherwise, it discards the oacket. When
`a repeater receives an SP, it checks whether it nas available storage,
`if it does, it makes one attempt to transmit a RSP and then discards
`it. When a repeater receives a RSP, it tests whether it needs one, if
`it does, it used the label, otherwise it discards it. The repeater
`currently simulated has buffer storage for two packets r ->ne exclu-
`sively used for packets directed towards the station, and the second
`for packets toward the terminal. In addition, the repeater can in-
`spect all packets that is receives, which are stored in common arrays
`in the simulation program. Thus, from a practical viewpoint, buffer
`storage for three packets per repeater are provided in the simulation
`program.
`
`7.28
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 10 of 201
`
`Netuovk Analysis Corporation
`
`>UV
`
`1 ) X □ E-
`
`U B
`H
`o
`H
`■ <
`a.
`O
`X
`M
`
`t
`
`K M M
`O D
`a |
`O O
`f I H
`
`2
`
`en
`a-
`H
`
`Ü
`
`I O
`
`E-"
`CO
`
`ro
`
`"^s
`
`7.29
`
`^^^a^MaaBa J
`
`
`
`n^
`
`h mm in- .1 »i-
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 11 of 201
`
`Network Analysis Corporation
`
`D
`
`Station
`
`The storage organization in the simulated station is shown
`in Figure 3.* There are two queues for active packets. Packets in
`these queues are active in the sense that they are retransmitted after
`short random periods of waiting until an Echo Ack is received. The
`active queue for long packets contains IP's from the station to ter-
`minals. Once an IP is Echo acknowledged, it is stored in the passive
`queue for a longer period, after which it is reactivated if an ETE
`Ack from the terminal is not received. The active queue for short
`packets contains ETE Acknowldegement packets to terminals, and these
`have priority over the long active apckets. The ETE Ack packets are,
`obviously, discarded once an Echo Ack is received. The point-to-point
`(PTP) network queue simulates the interaction of the packet radio net-
`work with a PTP network. When a new IP is received from a terminal,
`it is stored in the PTP network queue fir a random time, after which
`a response message containing several response IP's to that terminal
`are generated and placed into the active queue for long packets. The
`station responds immediately to SP's, and ignores RSP's.
`
`E. Flow Dia Trams of Devices
`
`Figures 4, 5, and 6 show the flow diagrams of the devices
`used in the simulator. These diagrams are simplified to the extent
`that they show "what to do" but not sufficiently detailed to show
`
`"How to do it." The latter depends on the particular system simulated,
`i.e., the routing, the channel configuration, etc.
`
`A device is called from the subroutine EVENT; the calling
`sequence includes, among others, an interrupt number which indicates
`to the device, the task that is has to perform. The only event which
`
`L'iSSf«!!!11?!?«.!-^ ^ storac?e of ip,s W* ETE Acks for transmission
`to terminals «« the packet radio network.
`
`7.30
`
` .
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 12 of 201
`
`I ■■!"■* W.#«MPI
`
`Network Analysis Corpomiion
`
`is external to a device Is that with an interrupt . 1. All other
`ThuTetSHt0 a f"" are dUe t0 eVentS generated ^ tte ^i« its-"-
`Thus, the number of exogeneous events is very small compared to the
`number of events generated by devices.- in particular, „hen the offered
`data rate to the system is high.
`
`hioh VS CXS" that When ^ """^ tra£flC "te t0 tte astern is
`h there wUl be many collisions of packets. Thus, to save com-
`
`con ntT^"01 eVi0e WaS "^ SO that " d0eS "<* —^ ^e
`content of the packet or whether the packet is addressed to it, at the
`
`oT::::::of pTet reception (see diagram=,•This is ^"the -d
`
`of packet reception, providing there was no interference
`There are parts of the subroutines of Repeater, station, and
`Termrnal wh^h are identical and thus, coded into subroutines. These
`relate to the identification of the header content, and the channe
`access mode. Some of these can be associated with the modem of the
`Physrcal devices. The parts which differ involve mainly the proces-
`
`IZ™ LT. T T" haS been identi£ied' e-9- **-** « ^e
`
`station, ETE Ack by end devices, etc.
`
`7.31
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 13 of 201
`
`Network Analysis Cor-poration
`
`1
`
`<
`
`u S s 0
`
`H
`
`q
`
`2
`O 9
`h
`U
`U
`M
`>
`Q
`
`i
`
`7.32
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 14 of 201
`
`Network Analysis Corporation
`
`r\
`c u
`0
`m 06
`v-y
`
`7.33
`
`■^MMMMÜ^^MM
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 15 of 201
`
`wmmrn
`
`n'»1 '" ' ' ■
`
`m m
`
`Network Analysis Corporation
`
`l^\
`
`c
`C
`3
`4J
`i |
`K-J
`
`2
`O
`
`u
`u
`
`Q
`
`u
`•^r
`
`MJ
`
`D
`0
`H
`h
`
`7.34
`
`t ——>^-
`
`
`
`■■■»^^^MW^OTH
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 16 of 201
`
`TERMINAL DEVICE FLOW DIAGRAM
`
`Network Analysis Corporation
`
`IP
`
`Generate
`Event for
`VPacket Reac-
`tivation
`
`Set
`ETE ACK
`
`Store Packet
`Set
`Parameters
`
`I
`
`Schedule
`Packet for
`rransiiissior
`
`I Return )
`
`FIGURE 4D
`7.35
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 17 of 201
`
`Network Analysis Corporation
`
`r\
`
`vv
`
`<
`in
`
`W
`
`e
`
`L
`
`7.36
`
`- — - ■ ■ - J
`
`
`
`BB^^n^nv^V
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 18 of 201
`
`Network Analysis Corporation
`
`IP, ETE ACK
`
`Increment
`Number of
`Transmission
`
`RSP, ECHO ACK
`
`ETE ACK
`
`/ReschedulA
`/Packet for \
`Vrransmissio/
`
`Schedule
`All Packets)
`^for Trans-
`
`I /Reschedules
`I
`
`Packet for
`fransmissior
`
`RETURN / y
`(
`
`FIGURE 5B
`
`STATION DEVICE FLOW DIAGRAM
`
`7.37
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 19 of 201
`
`-^ —-■--■• -■
`
`Network Analysio Corporation
`
`'Gcneratf'
`^Event end
`kof Receive
`
`FIGUHg 5C
`
`STATION DEVICE FLOW DIAGRAM
`
`7.38
`
`
`
`<mmmmmmmmmmma** ■
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 20 of 201
`
`Network Analysis Corporation
`
`Y
`-4-
`
`Select Packe
`with Highest
`Priority
`
`'Reschedule
`fPacket for
`VTransmissioi
`
`Transmit
`Packet
`
`(
`
`i
`I General
`
`Event for
`fn. of Trans
`•^mission
`
`r
`
`f Return J
`
`FIGURE 5D
`
`STATION DEVICE FLOW DIAGRAM
`
`7.39
`
`■
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 21 of 201
`
`■—""—— I" "'•■
`
`STATION DEVICE FLOW DIAGRAM
`
`Nctuurk At.nlyoia Corporation
`
`SP
`
`Droj- Packet
`Acknowledqec
`
`/Generate \
`fevent for \
`
`Transmit
`'ECHO ACK
`Packet
`
`I
`
`/Gunerat'
`Event fo:
`U-'nd of Iranj
`\_
`
`Schedule
`lacket for
`rransmi.ssioE
`
`Transmit
`RSP Packet
`
`I
`
`Generate
`Event for
`ind of Trans
`mission /
`
`C RETURN
`
`)
`FIGURE 5E
`7.40
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 22 of 201
`
`^ '
`
`Network Analysis Corporation
`
`>
`-H
`0)
`
`U I
`
`c
`■H
`
`I
`
`Q
`
`Q
`
`(9
`
`3
`O
`
`w
`u
`H
`>
`w
`Q
`a
`w
`EH
`<
`
`w
`a;
`
`7.41
`
`
`
`^■^^M
`
`wm^mnm^mm
`
`^ TIWIIPBP^'^'^w-w^'
`
`'"■ ■
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 23 of 201
`
`REPEATER DEVICE FLOW DIAGRAM
`
`Network Analysic Corporation
`
`IP,ETE ACK
`
`Drop Packet
`
`Acknowledged
`
`^Generate
`Svent end c-
`Receive
`
`Store Packet
`Set Para-
`
`^Schodule
`Packet for
`
`meters I
`fransmissior m
`
`f Return J
`
`riGURL 6B
`
`7.42
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 24 of 201
`
`Netuork Analysis Corporation
`
`r> u
`- i ■ U
`H w b
`
`4J
`Ck 0)
`o A:
`Q m
`
`7.43
`
`<
`
`3
`O
`
`M
`
`Q
`
`u u
`> u
`a u
`EH < u
`a,
`lü
`
`
`
`^'..J11 immmmm^p^i
`
`wpw^^^^w^w-
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 25 of 201
`
`Network Analysis Cor^joration
`
`VIII
`
`SUBROUTINES OF THE SIMULATOR
`
`A. Data Structure and Management Subroutines
`
`EVENT Takes the next event out of the Event Data Structure for
`
`execution.
`
`INHEAP Adds an event to the heap in the Event Data Structure.
`
`INMESS Allows the introduction of special messages such as ack-
`nowledgements, control messages and the like into the
`Message Data Structure.
`
`INPUT Reads the input parameters and determines the placing of
`repeaters and stations.
`
`MESSP.EL Is called by device routines to release a message as soon
`as all packets representing the message are deleted.
`
`NEWMESS Generates next exogenous message and adds to the Message
`Data Structure.
`
`NEWPACK Adds a new packet to the Packet Data Structure.
`
`NXTEVNT Adds a new event to the Event Data Structure.
`
`OUT
`
`Prints out intermediate data for debugging.
`
`OUTHEAP Takes the index of the next event time from the heap.
`
`PRSIM
`
`The driver routine. (main program)
`
`7.44
`
`„M—^a—aMa^a--.-
`
` 1
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 26 of 201
`
`— - —-—
`
`Network Analysis Corporation
`
`MEASURE Collects data on system performance.
`
`MCOUNT counts the number of packets associated with each terminal
`which are stored in the system.
`
`B. Communication and Device Subroutines
`
`REPEAT Main subroutine of repeater.
`
`STATION Main subroutine of station.
`
`TERMINAL Main subroutine of terminal.
`
`DEVINIT Reads parameters which devine the particular communication
`system, labels, and flow control parameters. Initializes
`states of devices.
`
`BGNPCV Maintains states of devices related to the RF channel (e.g.,
`number of overlapping packets).
`
`ENDRCV Same as above at the end of packet reception.
`
`ECHO
`
`SRER
`
`Records that a device is receiving an echo acknowledgement.
`
`Called when a non-overlapping packet is received for testing
`packet type and label.
`
`ALTROUT Called after repeater receives an RTS, checks whether
`repeater needs one, and has not used one before.
`
`SRTTRT Transmits packet and generates an event for the end of
`packet transmission.
`
`7.45
`
`
`
`r
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 27 of 201
`
`Network Analysis Corporation
`
`REPNEXT Determines which packet of a repeater is to be transmitted
`next.
`
`TERSTOR Stores correct IP's received by terminal and qene.-ates
`event for transmission of an ETE Ack.
`
`SNREEKO Called by station after receiving an Echo Ack. Identifies
`and maintains the queue in which the acknowledged packet is
`stored. If it is in IP, then it transfers the packet to
`another aueue where it waits for an ETE Ack or for reacti-
`vation.
`
`SHIFTQ Shifts packets in the various queues of the station.
`
`SNREPAK Called by station after correctly receiving a packet. If
`the packet is an ETE Ack, then subroutine drops the packets
`acknowledged, maintains proper queues and the message counts,
`If it is an IP, subroutine verifies that same packet has not
`been received before, and if so, it generates packet and an
`event for transmission of an ETE Ack, and also generates a
`random time and an event for the arrival of the response
`message from the PTP network.
`
`RESPONS Called by station, sets all response packets to a terminal
`into the active queue and generates events for transmitting
`them.
`
`SEKOTRT
`
`Used by station to transmit an Echo Ack for the last hoo,
`
`CONNECT Determines the most efficient repeater to which station
`should address packet when transmitting to a terminal.
`
`7.46
`
`
`
`- ■- ■ -- - • ■ —
`
`MÜ
`
`
`
`^^w»^www»»
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 28 of 201
`
`Network Analysis Corporation
`
`TRANSMT Called by a device which transmits a packet. Puts packet ]
`in list structure; determines all the devices that should
`receive the packet, ^he exact time for beginning to receive
`it; and generates the events to devices.
`
`C. Summary of Acronyms
`
`Ack - Acknowledgement
`
`AR - Active Receive
`
`ART - Active Receive and Transmit
`
`AT - Activ-i Transmit
`
`ETE - End-to-end
`
`HBH - Hop-by-hop
`
`IP ~ Information packet
`
`Label- An address assigned to a device for routine purnoses
`
`MHN - Maximum handover number
`
`MNT - Maximum number of transmissions
`
`Rp - Passive Receive
`
`PTP - Point-to-point
`
`RSP - Response to search packet
`
`SP - Search packet
`
`7.47
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 29 of 201
`Network Analysis Corporation
`
`>iiiii«pir»'iwMr«
`
`IX- OBSERVATION OF TRAFFIC FLOW IN THE PACKET RADIO NETWORK
`
`The first system simulated was a Common Channel Single Data Rate
`system, in which -He station is routing traffic as a repeater (Waive
`Station). We denote the system as CCSDR (NS). The system defined
`
`has a single data signalling rate for communication between terminal
`and repeater (or station) and in the repeater-station network; the
`channel is used in a half duplex mode. When the station is routing
`traffic as a repeater, it cannot receive packets not specifically
`addressed to it.
`
`In all experiments r^orted here, the labels of repeaters and
`staiton were preassigned. The hierarchical (directed) labelling
`scheme of the system in uhis experiment are shown in Figure 7. Fig-
`ure 8 shows the connectivity of the repeaters and station. That is,
`when a device transmits, all the devices connected to it by line are
`within an effective range and "hear" the transmission.
`
`The objective of the first series of experiments was to observe
`the detailed operation of devices and the efficiency of the system.
`The following observations were made:
`
`1. The "critical hop" in the system is that between the
`first level repeaters and the station. This was concluded
`by observing the frequency at which repeaters begin to search
`and at which they discarded packets, and from the obser-
`vation that there is no significant difference in the delay
`when the number of hops from the station that a packet travels
`is increased.
`
`7.48
`
`■■■■-■-■
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 30 of 201
`
`Network Analysis Corporation
`
`2. There is a higher probability of end-to-end successful
`completion when routing fron, the station to a terminal than
`when routing from a terminal to the station. F.actically,
`there is almost "no" difference in time delay between the
`delay of an information packet '.rom the terminal arriving at
`the station and the time that the terminal receives an ETE
`Aok from the station.
`
`3. Many packets associated with terminals that have de-
`parted from the system are routed in the network.
`
`can HThe Üfr* ^ imprOVi^ the routi"9 capaciblities of the station
`8 that "^^ °b5erVed- In P^ticular, one can see in Figures 7 and
`aben Tf «—«"»"» station is , there are only 4 repeaters
`labelled from the station. Consequently, the station is busy of the
`tim with non-useful traffic. This situation can be improved by
`changing the routing of the station so that: (1) it receives any
`packet that it can hear and wh .ch is (eventually) addressed to it•
`and 21 it transmits response packets to the repeater nearest to the
`termxnal along the routing path that it can reach. This change was
`implemented for all system studies subsequent to the initial experi-
`ments. Apart from the change implemented, the observation suggests
`that particular attention should be given to the design of the re-
`peater network in the neighborhood of the stations, it is also noted
`that these repeaters have a higher power duty cycle since they handle
`all p. ckets collected from other parts .e network. The routing
`change rr.de at the station enables the allocation of many more re-
`peaters in the neighborhoos of the station, than are functionally
`needed, without resulting an increase in the artivioial traffic gen-
`erated. The exact labelling of these repeaters is also not critical
`
`7.49
`
`■■ ■—-■ ■ ■ —
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 31 of 201
`
`Neiuork Anulyoic Corporation
`
`One of the reasons leading to observation 2 is that the station has
`a hlqhor probability then the first level repeaters of successful
`transmission over the critical hop, because it is the largest user
`and does not interfere with its own transmissions. Theoreticallv
`one may expect a similar conclusion when considering transmission'
`In a section of the network in which two repeaters, one of which
`"homes" on the other, compete. This, however may not be realized
`In the system simulated because of the limited storage available in
`repeaters.
`
`Observations 2 and 3 suggest a change in the Terminal-Station
`protocol. The basic question is whether a terminal should release
`itself from the system or whether it should be released by the .tatior
`The former was initially simulated. it was observed that in many
`cases, a terminal departed from the system after receiving an Echo
`to the ETE Ack for the last IP without this ETE Ack arriving at the
`station. This resulted in the reactivation if IP's bv the station
`for this terminal, the routing of these packets in the net, and then
`the maximum number of transmissions and search by the repeater nearest
`to the terminal. The protocol simulated m the systems discussed
`later is such that the last packet must always be from the station to
`the terminal. This transmission may be considered as a terminal re-
`lease packet. Another change in protocol implemented is that when-
`ever possible, the terminal acknowledges a sequence of packets rather
`than individual ones, to reduce the overhead in the direction towards
`the station.
`
`7.50
`
`■ ■ - --■■ — ■■ ■ -- ■
`
`
`
`■— »■
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 32 of 201
`
`Netwo) k Analysis Corporation
`
`F::GURE 7
`
`HIERARCHIAL LABELLING SCHEME
`
`7.51
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 33 of 201
`
`Network Analyeia Corp* vai ' <
`
`FIGURE 8
`
`CONNECTIVITY OF REPEATERS & STATIONS
`
`7.52
`
`J
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 34 of 201
`
`Network Analysis Cop'jorativn
`
`X. THE TRADEOFF BETWEEN TRANSMISSION RANGE OF DEVICES
`AND NETWORK INTERFERENCE
`
`For the experiments discusseo in the previous section, it was
`assumed that Repeater-Repeater range is the same as Terminal-Repeater
`range. This, however, is not always a realistic assumption since
`repeater.- can he placed on elevated areas and can have more power
`then terminals, {especially hand held terminals). Thus, if re-
`peaters are allocated for area coverage of terminals, the repeater
`range will be higher than terminal range and higher network con-
`nectivity or device interference will result.
`The problem which then arrises is to determine the impact of
`this interference on system performance. Alternatively, one may
`seek to reduce repeater transmission power when transmitting in
`the repeater-station network. To study this issue, two CCSDR systems
`were simulated, one with high interference CCSDR (HI), and the other
`with Low Interference CCSDR (LI). The routing labels of the two
`systmes were the same and are shown in Figure 7. Tha interference
`of the CCSDR (LI) system is shown in Figure 8 and the interference
`of the CCSDR (HI) system in Figure 9. Figure 9 shows only the
`connectivity of two devices in the network.
`The results are shown in Figure 10 and Table 1. Figure 10
`shows the throughput of the two systems as a function of time
`vhile Table 1 summarizes all other measures of performance. The
`third row of Table 1 summarizes performance of the high interference
`system under an improved set of repeater labels. This experiment
`is discussed in detail in the next section. It is clear that the
`high interference system is much better than the low interference
`system. The only measure of the low interference system which is
`better is terminal blocking which is a direct result of the low
`interference feature. In fact, CCSDR (LI) is saturated at the
`
`7.53
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 35 of 201
`
`Network Analysis Cürjor'ation
`
`offered traffic rate. This can be seon from the fact that the
`throuqhput is decreasinq as a function of time; the relatively
`hiqh total loss; and the low station response*. The CC.TDR (HI)
`with improved labels compared in Table 1, is better than the
`other two systems. This demonstrates the importance of proper
`
`labollinq. The experiments of this section demonstrate that it
`is preferable to use hiuh transmitter power to obtain long re-
`peater ranqe, despite the network interference that it results.
`
`* The average number station response packets assumed for these
`studios is 2.0.
`
`7.54
`
`
`
`r
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 36 of 201
`
`i ins*
`
`Network Analysis Corporation
`
`$
`
`FIGURE 9
`
`INTERFERENCE. OF CCSDR (HI) SYSTEM
`
`7.55
`
`
`
`a:
`.-- w
`u
`u
`
`(A
`
`<
`
`Cd
`E-
`
`>
`
`S
`=
`5
`c
`B
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 37 of 201
`
`•^—**^^^m**
`
`Netuofk Analyaie Coipcvution
`
`Mt
`
`S ^
`
`a
`u
`U
`
`c h
`o c
`LP 1-5
`
`M
`
`D
`C
`H
`
`o
`o
`o
`
`O
`o
`un
`
`THROUGHPUT [%]
`
`vo
`
`ci
`
`7.56
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 38 of 201
`
`I"""1 ' " ■"
`
`Network Analyeie Coppi ration
`
`• i H
`
`M M B i u w
`
`Ei B
`
`.J a.
`■t M
`p C
`
`1 b ^
`
`tfl
`
`►J
`
`Bi
`
`3
`M 2
`k □ o 3
`
`J» 3)
`
`.:
`~
`• M
`i £
`5 H >•
`s W a
`
`w
`u. • r
`o 6
`t/)
`M g
`H H OH
`P < to
`2
`Ö
`
`r-<
`
`-t
`
`o
`
`m
`' ^ c !■ ~l
`o H
`H
`0
`H
`Cl
`M a. H H '/j
`a M 0)
`
`E< p
`
`t.5
`
`U W
`X;. A
`
`3
`S
`
`o
`
`»0
`
`o
`in
`
`o
`
`r i
`
`a
`01
`
`.'1
`CO
`
`co
`(Ti
`
`o
`
`n
`0>
`
`in
`m
`
`I
`
`— >
`0
`u. u
`Q a.
`to E
`U M
`U —
`
`a;
`D
`t/j u
`u
`
`a
`w
`o
`0
`
`7.57
`
`
`
`Case 1:16-cv-02690-AT Document 121-13 Filed 08/05/16 Page 39 of 201
`
`Netuurk Analysis Corporation
`
`XI. SINGLE VERSUS DUAL DATA SIGNALLING RATES NETWORKS
`
`The results cf the previous section demonstrate that a better
`performance system is obtained when repeaters and station use hiqh
`power to obtain long range despite the interference that results.
`We nov; examine the problem of whether repeaters and station should
`use their fixed power budget
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
