(12) United States Patent
`Panasik et al.
`
`I IIIII IIIIIIII Ill lllll lllll lllll lllll lllll lllll lllll lllll 111111111111111111
`US006643278Bl
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 6,643,278 Bl
`Nov. 4, 2003
`
`(54) WIRELESS NETWORK CIRCUITS,
`SYSTEMS, AND METHODS FOR
`FREQUENCY HOPPING WITH REDUCED
`PACKET INTERFERENCE
`
`(75)
`
`Inventors: Carl M. Panasik, Garland, TX (US);
`Thomas M. Siep, Garland, TX (US)
`
`(73) Assignee: Texas Instruments Incorporated,
`Dallas, TX (US)
`
`( *) Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by O days.
`
`(21) Appl. No.: 09/473,337
`
`(22)
`
`Filed:
`
`Dec. 28, 1999
`
`(60)
`
`(51)
`(52)
`(58)
`
`Related U.S. Application Data
`Provisional application No. 60/125,573, filed on Mar. 23,
`1999.
`
`Int. Cl.7 .................................................. H04Q 7/00
`U.S. Cl. ........................................ 370/330; 370/344
`Field of Search ................................. 370/330, 480,
`370/342, 441, 442, 241, 252, 254, 389,
`392, 465, 394, 344, 345; 375/130, 131,
`132-134; 380/49, 9, 48, 59; 455/422, 436
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`4,716,573 A
`5,506,863 A *
`5,528,622 A
`5,778,075 A *
`5,809,059 A
`
`12/1987
`4/1996
`6/1996
`7/1998
`9/1998
`
`Bergstrom et al.
`Meidan et al. .............. 375/134
`Cadd et al.
`Haartsen ..................... 375/138
`Souissi et al.
`
`* cited by examiner
`
`Primary Examiner-Dang Ton
`(74) Attorney, Agent, or Firm-Ronald 0. Neerings; Wade
`James Brady, III; Frederick J. Telecky, Jr.
`
`(57)
`
`ABSTRACT
`
`A method (10) for determining a frequency hopping
`sequence for a newly-entering network. The method com(cid:173)
`prises the step of scanning (16) a plurality of frequency
`channels. For each of the plurality of frequency channels,
`the scanning step comprises detecting whether a signal (18,
`22) exists on the channel and recording information (20, 24)
`corresponding to each channel on which a signal is detected.
`Finally, and responsive to the recorded information, the
`method forms (30) the frequency hopping sequence.
`
`25 Claims, 2 Drawing Sheets
`
`12
`
`COMMENCE NEW HOPPING
`SEQUENCE DETERMINATION: START
`UP NEWLY-ENTERING NETWORK
`
`14
`
`DETERMINE FIRST CHANNEL FOR ANALYSIS
`
`10
`
`\
`
`16
`
`SCAN SELECTED CHANNEL
`FOR EXISTING SIGNAL
`
`20
`
`RECORD TIME SLOT
`AND CHANNEL
`
`24
`
`RECORD USAGE
`CHARACTERISTICS
`
`28
`
`NEXT CHANNEL
`
`NO
`
`CREATE HOPPING SEQUENCE
`FOR NEWLY-ENTERING NETWORK
`
`MODIFY NEWLY-CREATED HOPPING
`SEQUENCE TO AVOID FIXED
`INTERFERENCE
`
`30
`
`32
`
`YES
`
`WAIT AT LEAST TWO SLOTS
`AFTER
`INCUMBENT USES FIRST
`CHANNEL
`IN ITS SEQUENCE
`
`38
`
`START FREQUENCY HOPPING
`
`36
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0001
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`U.S. Patent
`
`Nov. 4, 2003
`
`Sheet 1 of 2
`
`US 6,643,278 Bl
`
`~!
`~ !c2
`~ 1
`
`I
`I
`I
`I
`:
`
`......
`
`.............. -
`
`I
`I
`I
`I
`I
`: P12 :
`
`I
`I
`
`1
`I
`I
`I
`
`I
`I
`I P2l : C7
`
`,
`I
`
`I
`
`,
`I
`
`I
`
`1
`I
`I
`I
`
`P17
`
`::m/
`~ P25I
`
`I P16
`I
`
`I
`
`I
`
`I
`
`I
`
`I
`I
`I
`I
`1 P21 1
`
`I ~!
`:~:
`
`I
`1 P11
`
`I
`I
`
`I
`
`I
`
`-
`
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`ta
`
`FREQUENCY
`CHANNEL
`
`40
`"\
`
`42
`\
`
`'[_ - RADIO
`
`LEGEND
`~ FIRST NETWORK
`PACKET, P1 n
`
`~ SECOND NETWORK
`PACKET, P2n
`
`FIXED
`lZSZSJ INTERFERENCE (Fl)
`
`I
`t7
`
`I
`I
`I P241
`I
`I
`I
`I
`"
`I
`I
`I
`I C 1 I
`I
`I
`I
`I
`I
`I
`I
`I C4
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`I
`t 1
`t2 t3
`
`I
`
`I
`I
`
`I
`I
`I
`t5 t5
`t4
`FIG. 1
`
`48~~ ~50
`
`52
`6 ~ ....
`
`HOST
`1/F
`
`MAC
`CONTROLLER "'-46
`
`44
`\
`
`PHYSICAL
`ENGINE
`
`~
`
`-
`
`-
`
`RSSI
`
`-
`
`PREAMBLE +
`DATA
`-
`
`-
`
`TX/RX
`FREQUENCY
`
`SUBCHANNEL
`SCAN
`
`HOP
`SEQUENCE
`HOP
`CONTROL
`SCAN
`COMMAND
`
`-
`
`DATA
`SCAN
`RESULTS
`
`FIG. 3
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0002
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`U.S. Patent
`
`Nov. 4, 2003
`
`Sheet 2 of 2
`
`US 6,643,278 Bl
`
`FIG. 2
`
`10
`\
`
`12
`
`14
`
`16
`
`COMMENCE NEW HOPPING
`SEQUENCE DETERMINATION: START
`UP NEWLY-ENTERING NETWORK
`
`SCAN SELECTED CHANNEL
`FOR EXISTING SIGNAL
`
`YES
`
`RECORD TIME SLOT
`AND CHANNEL
`
`24
`
`YES
`
`RECORD USAGE
`CHARACTER I ST! CS
`
`28
`
`NEXT CHANNEL
`
`NO
`
`CREATE HOPPING SEQUENCE
`FOR NEWLY-ENTER[NG NETWORK
`
`MODIFY NEWLY-CREATED HOPPING
`SEQUENCE TO AVOID FIXED INTERFERENCE
`
`30
`
`32
`
`YES
`
`WAIT AT LEAST TWO SLOTS
`AFTER INCUMBENT USES FIRST
`CHANNEL IN
`ITS SEQUENCE
`
`NO
`
`38
`
`START FREQUENCY HOPPING
`
`36
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0003
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`US 6,643,278 Bl
`
`1
`WIRELESS NETWORK CIRCUITS,
`SYSTEMS, AND METHODS FOR
`FREQUENCY HOPPING WITH REDUCED
`PACKET INTERFERENCE
`
`This application claims the benefit of Provisional Appli(cid:173)
`cation Ser. No. 60/125,573 filed Mar. 23, 1999.
`
`CROSS-REFERENCES TO RELATED
`APPLICATIONS
`
`Not Applicable.
`
`STATEMENT REGARDING FEDERALLY
`SPONSORED RESEARCH OR DEVELOPMENT
`
`Not Applicable.
`
`BACKGROUND OF THE INVENTION
`
`The present embodiments relate to wireless communica(cid:173)
`tion systems, and are more particularly directed to such
`systems using frequency hopping.
`Wireless networks are becoming increasingly popular,
`and in this regard there has been improvement in many
`aspects of such networks. Some improvements relate to 25
`configurations that permit simultaneously operation of dif(cid:173)
`ferent networks where there is minimal or no interference
`between communications belonging to each of the networks.
`In this respect, the term network is used, and is further used
`in the same manner for the remainder of this document, to 30
`describe a system consisting of an organized group of
`intercommunicating devices. Further in this respect, the
`different networks may be labeled according to a first
`network that is already transmitting in time followed by a
`second network in time seeking to transmit and thereby 35
`possibly communicating and causing interference due to a
`communication overlapping the pre-existing communica(cid:173)
`tion of the first network. Accordingly, to facilitate the
`remaining discussion, such a first network is referred to as
`an incumbent network, while the network which seeks to 40
`communicate, or in fact does communicate, after the incum(cid:173)
`bent network is referred to as the newly-entering network.
`Given this terminology, the present background and embodi(cid:173)
`ments discussed below are directed to reducing interference
`between incumbent network communications and newly- 45
`entering network communications.
`One approach to reducing the above-introduced interfer(cid:173)
`ence is known in the art as spread spectrum frequency
`hopping and is sometimes referred to more simply as
`frequency hopping. In frequency hopping, a newly-entering 50
`network transmitter transmits packets of information at
`different frequencies in an effort to reduce the chance that
`the packet will interfere or "collide" with a packet transmit(cid:173)
`ted at a frequency by a transmitter in an incumbent network.
`The change between frequencies, that is, from one frequency 55
`to another, is said to be a "hop" between the frequencies.
`Moreover, the goal is such that each packet from a newly(cid:173)
`entering network is transmitted at a frequency which neither
`overlaps nor is near enough to a frequency at which an
`incumbent network is transmitting. Further in this regard, 60
`some systems (e.g., using Bluetooth protocol) transmit each
`successive packet at a different frequency, that is, the trans(cid:173)
`mitter is "hopping" to a different frequency for each packet.
`Alternatively, others systems (e.g., IEEE 802.11) transmit a
`first set of packets at a first frequency, and then hop to a 65
`second frequency to transmit a second set of packets, and so
`forth for numerous different sets of packets at numerous
`
`5
`
`2
`different respective frequencies. Note further that if inter(cid:173)
`ference or a collision does occur, it typically corrupts the
`data of both packets, that is, the data transmitted by both the
`newly-entering network and the incumbent network. As a
`result, both networks are then required to re-transmit the
`packets an additional time so as to replace the corrupted data
`resulting from the collision.
`In an effort to achieve minimal packet collision using
`frequency hopping, two prior art methods have arisen for
`10 determining the different frequencies to which a network
`will hop. In a first method, a frequency hopping network
`uses a pre-ordained hopping sequence. This first approach is
`used by way of example under the IEEE 802.11 standard. In
`a second method, a seed is provided to a pseudo-random
`15 generator which produces a corresponding pseudo-random
`series of frequencies along which the network hops. This
`second approach is used by way of example under the fairly
`recently developed Bluetooth protocol. Both of these
`approaches have achieved some level of success in reducing
`20 the amount of inter-network packet collision. Nevertheless,
`the present inventors have empirically determined that by
`locating two or more different networks in the same vicinity
`such that transmissions from each different network effec-
`tively compete for airtime, there still arises a considerable
`amount of packet collisions, thereby reducing the effective
`transmission rate for each network.
`Frequency hopping as described thus far reduces the
`chances of interference between a packet from newly(cid:173)
`entering network and a packet from an incumbent network.
`Further in this regard and by way of additional background,
`FIG. 1 illustrates communications of such packets and, as
`detailed below, it also illustrates instances where packet
`collisions occur. Looking to FIG. 1 in greater detail, its
`horizontal axis illustrates time ( or time slots), and its vertical
`axis indicates frequency. Additionally, FIG. 1 illustrates a
`number of blocks, where each block is intended to depict a
`packet as transmitted by either an incumbent network or a
`newly-entering network. Further in this regard, note that the
`term "packet" is used in this document to define a block of
`information sent in a finite period of time, where subsequent
`such packets are sent at other times. This block of informa-
`tion may take on various forms, and sometimes includes
`different information types such as a preamble or other type
`of control information, followed by user information which
`is sometimes also referred to as user data. Further, the
`overall packet also may be referred to in the art by other
`names, such as a frame, and thus these other information
`blocks are also intended as included within the term
`"packet" for purposes of defining the present inventive
`scope. In any event, returning to FIG. 1, for the sake of
`reference, each packet illustrated in FIG. 1 is labeled with an
`identifier using the letter "P" (i.e., for packet) and following
`after that letter is a number corresponding to the network
`which transmitted the packet. More particularly, packets
`transmitted by the first network (i.e., the incumbent
`network) are labeled with an identifier Pl while packets
`transmitted by the second network (i.e., the newly-entering
`network) are labeled with an identifier P2. Further, the
`subscript for each packet identifies a time period encom(cid:173)
`passed by the duration of the packet. For example, during a
`time to, the first network transmits a packet Pl0 while also
`during time t0 the second network transmits a packet P20 .
`Further in this regard, in the prior art transmissions by the
`first network are asynchronous with respect to transmissions
`of the second network, both in start time and periodicity.
`Thus, time t0 is only meant as a relative indication for the
`first packet from each network, and it is not intended to
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0004
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`US 6,643,278 Bl
`
`3
`suggest that the packets from both networks begin and end
`at the same time.
`With respect to all packets in FIG. 1, the preceding
`demonstrates that each packet begins at a certain time, ends
`at a later time, and fills a certain frequency range (where the
`range is referred to as a channel). As a result and as
`described below, interference may occur if the area in FIG.
`1 defined by a packet overlaps or is within a certain distance
`of a packet from another wireless link. Indeed and as
`discussed below, such interference may occur in one of four
`different ways.
`in FIG. 1 illustrates a first type of packet
`Time t1
`interference, where it may be seen that the first network
`transmits a packet Pl 1 After packet Pl 1 commences but also
`during time t1 the second network transmits a packet P21 .
`The overlap of packets Pl 1 and P2 1 is shown as a first
`collision C1 . Note that the horizontal alignment of packets
`Pl 1 and P2 1 graphically indicates that in the example of
`collision C1 , both packets occupy the same frequency chan(cid:173)
`nel. Thus, collision C1 represents an example where two
`different networks attempt to transmit packets during an
`overlapping time period and along the same channel.
`Before proceeding with other types of packet collisions,
`an additional discussion is noteworthy with respect to a
`methodology which has been used to further reduce the
`likelihood and impact of packet collisions such as collision
`C1 . More particularly, this additional methodology is
`referred to in the art as listen-before-talk ("LET"). In an
`LET system, the system uses the hopping sequence
`described above, but prior to transmitting along a channel in
`the sequence the system monitors ( or "listens") at the
`channel to determine if there is another packet already
`occupying that channel during the current time. Returning to
`packet Pl 1 by way of example, if the second network
`employed LET, then it would listen at the desired channel at
`which it intended to transmit P2 1 and would therefore detect
`the presence of packet Pl1 . As a result, the second network
`would avoid collision Cl by not transmitting packet P2 1 at
`the desired frequency, but instead it would delay a random
`period and then proceed to the next designated channel of its
`hopping sequence. Next, the second network would listen at
`that next designated channel to again determine if that
`channel was occupied by a packet from another network,
`and if no packet was detected then the second network
`would transmit its packet; however, if this next designated
`channel also was occupied, then the second network would
`continue to examine additional channels in this same manner
`until a channel was detected without being occupied by a
`packet from another network, at which time the second
`network would transmit its packet along the now unoccupied
`channel. Given this process, however, note that a delay
`arises in LET systems, where the amount of delay depends
`on the number of times that the LET network is forced to
`listen, detect, and advance from an occupied channel, and
`then delay an additional random period to listen, detect, and
`transmit along an unoccupied channel.
`While LET as shown above reduces the possibility of
`collisions, it also has drawbacks. For example, LET delays
`transmission by the network which was prepared to transmit
`along a channel but was prevented from doing so due to an
`already-transmitted packet in the desired channel. As
`another example, it adds an element of delay to each packet
`due to its listening aspect. Also, all the devices in an
`environment must utilize LET to gain the most benefit
`(fairness) of the scheme. As still another example, some
`protocols ( e.g., Bluetooth) utilized in the unlicensed bands
`do not support LET, while such protocols may nonetheless
`
`5
`
`4
`provide other beneficial aspects and, thus, the choice to use
`such a protocol is a tradeoff in that other aspects are obtained
`without the availability of LET.
`Time t2 in FIG. 1 illustrates a second type of packet
`interference in connection with a collision C2 occurring
`between a first network packet Pl2 and a second network
`packet P22 . For collision C2
`the incumbent first network
`transmits packet Pl2 during a period including time t2 and at
`a first channel, and thereafter the second network transmits
`10 packet P22 also during a period including time t2 (i.e., the
`periods of the packets overlap). Packet P22 is transmitted at
`a second channel which, while different than the channel of
`packet Pl2 , it is immediately adjacent the channel occupied
`by packet Pl2 . Further in this regard, it is known in the art
`15 that while packets occupy a certain channel as shown by the
`vertical displacement of a packet in FIG. 1, there is an
`additional tendency for a packet to provide slight interfer(cid:173)
`ence or "splatter" into adjacent frequency channels. As a
`result of this effect, even though packets Pl2 and P22 occupy
`20 different channels, they are still in adjacent channels and,
`thus, they are close enough to one another in frequency such
`that the splatter effect causes a collision between the packets.
`Indeed, in some networks the filters used are relatively
`inexpensive and, as a result, the concept illustrated with
`25 packets Pl2 and P2 2 may also apply to next-adjacent
`channels, that is, to the channels that are one more channel
`away from the channels adjacent to the channel in which a
`packet is transmitted. Thus, collision C2 represents an
`example where two different networks attempt to transmit
`30 packets during an overlapping time period and along adja(cid:173)
`cent ( or next-adjacent) frequency channels. Here, if neither
`network uses LET, then both packets Pl2 and P22 will
`require retransmission due to the collision. If, however, the
`network that intended to transmit the second packet of the
`35 two uses LET, then note first that LET mechanisms are less
`likely to correctly discern an adjacent channel collision.
`However, if the LET mechanism does recognize the poten(cid:173)
`tial adjacent channel collision, then the second packet is not
`transmitted along the channel represented by P22 and instead
`40 that packet is delayed. This delay, while diminishing the
`effective transmission of the second network, avoids any
`disturbance to the first already-existing packet. In the
`example of time t2
`therefore, if the second network uses
`LET, then packet Pl2 will not be disturbed because the
`45 second network will move the transmission of packet P22 to
`a different channel.
`Time t4 in FIG. 1 illustrates a third type of packet
`interference in connection with a collision C4 , which is
`comparable to collision C2 except that for collision C4 the
`50 networks transmit in opposite order. More particularly, for
`collision C4 , the second network first transmits a packet P2 4
`and, thereafter, the first network transmits a packet Pl4 . The
`duration of both of these packets overlaps time t4 , and again
`their channels are adjacent to one another rather than being
`55 the same channel. Nonetheless, the splatter effect again
`causes sufficient reach of each packet into the adjacent
`channel such that a collision occurs. Here, if neither network
`uses LET, then both packets P2 4 and Pl 4 require
`re-transmission due to the collision; if, however, the network
`60 transmitting the second packet in time (i.e., Pl 4 ) of the two
`which would otherwise collide uses LET, then only that
`packet is delayed and the first already-existing packet (i.e.,
`P24) is not disturbed.
`Time t7 in FIG. 1 illustrates a fourth type of packet
`65 interference in connection with a collision C7 , which is
`comparable to collision C1 except that for collision C7 the
`networks transmit in opposite order. More particularly, for
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0005
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`US 6,643,278 Bl
`
`5
`collision C7 , the second network first transmits a packet P27
`and, thereafter, the first network transmits a packet Pl7 . The
`duration of both of these packets overlap time t7 and their
`channels are the same. As a result, collision C7 occurs
`( assuming the last network to transmit, which here is the first 5
`network, does not use LET).
`FIG. 1 illustrates an additional type of potential interfer(cid:173)
`ence by depicting a band of fixed interference Fl. Fixed
`interference FI is intended to represent a non-network source
`of radio frequency transmission that remains at the same 10
`frequency for numerous time slots. Such fixed interference
`may arise from various devices, such as a leaking micro(cid:173)
`wave oven by way of example. In any event, note at time t5
`that the second network transmits a packet P25 , and the
`channel along which that packet is transmitted overlaps 15
`fixed interference FI. As a result, fixed interference FI
`interferes with packet P2 5 , thereby requiring it to be
`re-transmitted. Once more, however, if the second network
`were to implement LET, then assuming fixed interference FI
`were detected during the listening operation of the LET, then 20
`packet P25 would not be transmitted so as to avoid the
`otherwise imminent interference. Lastly, while the example
`of packet P25 demonstrates a data collision where the packet
`uses the same channel as the fixed interference, note further
`that fixed interference also may disturb packets in a channel 25
`that is adjacent to the channel including the fixed interfer(cid:173)
`ence. Once more, because some networks use relatively
`inexpensive filters, the fixed interference may corrupt pack-
`ets which are either in a channel which is immediately
`adjacent to the fixed interference or which are in the next- 30
`adjacent channel (i.e., a channel which is next to the channel
`that is immediately adjacent to the fixed interference).
`In view of the above, one skilled in the art should
`appreciate there are various opportunities for packet colli(cid:173)
`sion or packet interference to occur. Indeed, referring to FIG. 35
`1, the examples above demonstrate that an area may be
`described around each packet, where the packet is disturbed
`if another packet occurs within that area. Thus, this area,
`which may be perceived as a window or zone around the
`packet, is not only defined by the dimensions of the packet,
`but extends both before and after the packet by the width of
`another potentially-interfering packet, and extends above
`and below the packet channel through the height of at least
`the adjacent channel above and below the packet frequency
`channel. Still further, note that the packet sizes for both 45
`networks shown in FIG. 1 are the same size by way of
`example; however, in some contexts, an incumbent network
`may use packets of different dimension (i.e., either in
`frequency and/or time) relative to the newly-entering net(cid:173)
`work. In these cases, the packet size for the incumbent as 50
`well as the packet size for the newly-entering network, in
`addition to the window-affecting factors described above, all
`further define a two-dimensional area relative to a newly(cid:173)
`entering packet in which interference may occur. Given the
`size of the two-dimensional area, therefore, there remains a 55
`possibility of packet disturbance even given the pseudo
`random nature of hopping spread spectrum RF communi(cid:173)
`cations.
`As an additional consideration relative to avoiding packet
`collisions, it is further noted that the Federal Communica(cid:173)
`tions Commission ("FCC") imposes a restriction on the art
`in the Industrial Scientific Medical ("ISM") bands.
`Specifically, the FCC explicitly forbids independent net(cid:173)
`works to expressly cooperate in allocation of the wireless
`medium.
`In view of the above, there arises a need to reduce the
`possibility of packet collision and interference, and prefer-
`
`6
`ably to do so in a manner that may be used with protocols
`that do not support LET. The preferred embodiment
`addresses these goals, as is explored below. In addition,
`there arises a need to achieve the above goals while com(cid:173)
`plying the with the above-described FCC requirements. The
`preferred embodiments described below avoid these require(cid:173)
`ments by not requiring the newly entering network to have
`knowledge of or cooperation with the incumbent network.
`
`BRIEF SUMMARY OF THE INVENTION
`
`In the preferred embodiment, there is a method for
`determining a frequency hopping sequence for a newly(cid:173)
`entering network. The method comprises the step of scan(cid:173)
`ning a plurality of frequency channels. For each of the
`plurality of frequency channels, the scanning step comprises
`detecting whether a signal exists on the channel and record(cid:173)
`ing information corresponding to each channel on which a
`signal is detected. Finally, and responsive to the recorded
`information, the method forms the frequency hopping
`sequence. Other circuits, systems, and methods are also
`disclosed and claimed.
`
`BRIEF DESCRIPTION OF THE SEVERAL
`VIEWS OF THE DRAWING
`
`FIG. 1 illustrates various packets transmitted by a first and
`second network and demonstrates potential collisions
`between such packets as well as interference from a band of
`fixed interference.
`FIG. 2 illustrates a flow chart of the preferred embodiment
`as implemented in a method performed by a network trans(cid:173)
`ceiver.
`FIG. 3 illustrates a block diagram of a network transceiver
`operable to perform the method shown in FIG. 2.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`FIG. 1 was described in the Background Of The Invention
`40 section of this document and the reader is assumed familiar
`with that description.
`FIG. 2 illustrates a flow chart of a method 10 according
`to the preferred embodiment and for operating a wireless
`network so as to reduce the drawbacks described above in
`connection with the prior art. By way of introduction to this
`preferred embodiment, the following discussion demon-
`strates that by the conclusion of method 10 an improved
`hopping sequence is generated for a wireless network. The
`hopping sequence is improved in two respects. First, the
`hopping sequence is such that packets may be communi(cid:173)
`cated according to it and results in a reduced amount of
`packet collisions as compared to a prior art non-LET wire(cid:173)
`less frequency hopping system. Second, the hopping
`sequence is such that packets may be communicated accord(cid:173)
`ing to it and results in a reduced incidence of conflict with
`fixed interference as compared to a prior art non-LET
`wireless frequency hopping system. These benefits are illus(cid:173)
`trated in greater detail after the following detailed discussion
`of method 10. Finally, it should be noted that method 10 may
`60 be implemented in connection with various types of wireless
`networks as may be ascertained by one skilled in the art and
`as further addressed later. Additionally, such a person also
`may determine various different circuits and software imple(cid:173)
`mentations given the selected network, as is also explored
`65 later by way of example.
`Method 10 begins with a step 12 where the wireless
`network begins the determination of a new hopping
`
`Marvell Semiconductor, Inc. - Ex. 1015, Page 0006
`IPR2019-01350 (Marvell Semiconductor, Inc. v. Uniloc 2017 LLC)
`
`

`

`US 6,643,278 Bl
`
`7
`sequence to be used for intercommunications on the network
`(i.e., by all transmitters, receivers, and transceivers in the
`network). To facilitate the remaining discussion, the network
`which will use this new hopping sequence is referred to as
`the newly-entering network. This terminology is chosen 5
`because the newly-entering network's communications are
`new with respect to any one or more incumbent networks
`that already may be communicating along the frequency
`band to be used by the newly-entering network. In the
`preferred embodiment, step 12 occurs at network start-up, 10
`such as when a first transceiver of the newly-entering
`network is turned on or is otherwise initialized. Next,
`method 10 continues to step 14.
`In step 14, a first frequency channel is selected for
`analysis. More particularly and as will become apparent 15
`given the remaining discussion of method 10, in the pre(cid:173)
`ferred embodiment each channel along which the newly(cid:173)
`entering network may transmit is individually analyzed by
`method 10 at least once. Accordingly, step 14 operates so
`that a first one of these channels is selected to be analyzed. 20
`This selection may be implemented in various fashions, such
`as by assigning a unique and ascending number to each
`increasing frequency channel which is available to the
`newly-entering network, and then step 14 may operate by
`initializing a counter to the first assigned number. Other 25
`implementations may be ascertained by one skilled in the
`art. In any event, once a first channel is selected for analysis,
`method 10 continues to step 16.
`In step 16, the channel selected by step 14 is scanned to
`determine if there is an existing signal in that channel. In the 30
`preferred embodiment, the known receive signal strength
`indicator ("RSSI") is used as the scan technique. Note that
`an existing signal may be detected in the scanned channel
`due to various events as illustrated earlier in connection with
`FIG. 1. For example, an existing signal will be detected in 35
`step 16 if there is fixed interference in the scanned channel
`(or in a channel one or two adjacent channel locations from
`the scanned channel). As another example, an existing signal
`will be detected in step 16 if another network has transmitted
`a packet that, during the duration of the scan, is either in the 40
`scanned channel or in a channel that is adjacent the scanned
`channel. Each of these possibilities is responded to by one
`or more additional steps, as discussed below. Following the
`scan of step 16, method 10 continues to step 18.
`Step 18 directs the flow of method 10 if the interference, 45
`if any, detected in step 16 is fixed interference. The deter(cid:173)
`mination of whether a particular detected interference is
`fixed interference ( as opposed to packet interference) may
`be made in various fashions. A simple approach is to wait on
`an occupied channel for a period of time which exceeds all 50
`known packet lengths (0.4 seconds by FCC part 15 rules). In
`a faster and preferred approach, the instance of a fixed
`interferer is determined by determining its occupied
`bandwidth, which is very small relative to data carrying
`modulated signals. More particularly, many scan circuits are 55
`available which can be configured according to the preferred
`embodiment to determine the bandwidth of a received signal
`by stepping through several sub-channels of the particular
`channel. In the process, the scanning circuit collects the
`RSSI as a function of each sub-channel and determines the 60
`half-power points, which is the bandwidth. Thus, once all
`sub-channels for the scanned channel are evaluated, and
`assuming that interference is detected on at least one of those
`sub-channels, then it may be further determined that the
`interference is fixed interference based on the bandwidth 65
`identified across all sub-channels. Specifically, fixed inter(cid:173)
`ference typically occupies only ten percent or less of the
`
`8
`entire channel bandwidth; thus, if ten same-sized sub(cid:173)
`channels are scanned for a given evaluated channel, and if
`the interference detected is only in one or two of those
`sub-channels, then the preferred embodiment determines
`that the detected interference is fixed interference; to the
`contrary, if interference is detected across most or all of
`those sub-channels, then the preferred embodiment deter(cid:173)
`mines that the detected interference is packet interference. In
`any event, if fixed interference is detected, then step 18
`directs the flow to a step 20. To the contrary, if either no
`interference is detected, or if interference which is not fixed
`interference is detected, then method 10 continues from step
`18 to step 22. Each of these alternative paths is described
`below.
`In step 20, having been reached due to the detection of
`fixed interference existing in the scanned channel, method
`10 records an indication of the time slot and channel in
`which the fixed interference was detected. The use of this
`information is detailed later. Thereafter, method 10 contin(cid:173)
`ues from step 20 to step 26, which is discussed following the
`discussion below concerning steps 22 and 24.
`Step 22 directs the flow of method 10 if the potential
`interference, if any