as) United States
`a2) Patent Application Publication (10) Pub. No.: US 2005/0157675 Al
`(43) Pub. Date:
`Jul. 21, 2005
`Feder et al.
`
`
`US 20050157675A1
`
`(54) METHOD AND APPARATUS FOR
`CELLULAR COMMUNICATION OVER DATA
`NETWORKS
`
`(76)
`
`Inventors: Peretz Moshes Feder, Englewood, NJ
`(US); Jungsang Kim, Chapel Hill, NC
`(US); Zhengxiang Ma, Summit, NJ
`(US); Anatoli Olkhovets, Piscataway,
`NJ (US); Arnold B. Siegel, Alpine, NJ
`(US); Theodore Sizer I, Little Silver,
`NJ (US); Michael George Zierdt, Belle
`Mead, NJ (US)
`
`Correspondence Address:
`Lucent Technologies Inc.
`Docket Administrator
`Room 3J-219
`101 Crawfords Corner Road
`Holmdel, NJ 07733-3030 (US)
`
`(21)
`
`Appl. No.:
`
`10/884,203
`
`(22)
`
`Filed:
`
`Jun. 30, 2004
`
`Related U.S. Application Data
`
`(60)
`
`Provisional application No. 60/536,871, filed on Jan.
`16, 2004.
`
`219 :
`
`Publication Classification
`
`(51)
`(52)
`
`Unt. C17 caccccsosssssssssssnsstsntnessensenee H04Q 7/20
`US. Ch.
`cacsessesssssssssnsntsesntnetnsnatstnsiessee 370/328
`
`(57)
`
`ABSTRACT
`
`Cellular signals or other wireless signals/messages are intro-
`duced into a building or to an outside location by transmit-
`ting packets corresponding to those signals over a data
`network and low cost cables to designated locations within
`the data network. Once the designated packets containing
`the signals reach the destination, they are then broadcast
`over the air to a terminal capable of receiving the wireless
`message. In
`a
`first embodiment, an in-building gigabit
`Ethernet network, such as that currently existing presently in
`many buildings, is used to distribute radio signals indoors.
`Instead of transmitting the radio signals over the air from a
`repeater connected to a base station, coded baseband signals
`generated by the coding processor (e.g.,
`a CDMA Modem
`Unit) in the base station are packetized and sent over the
`Ethernet network to radio processing equipment and anten-
`nas distributed throughout the building. The radio process-
`ing equipment strips the packet headers from the baseband
`signal packets so those signals can be broadcast via the
`antennas to one or more mobile terminals.
`
`
`
`
`
`
`
`204 |- 5937
`
`
`
`
`
`
` {
`D)
`205
`
`206 \
`
`
`
`
`
`
`212
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Page 1
`
`
`
`
`
`CommScope Ex. 1018
`
`

`

`Patent Application Publication
`
`Jul. 21,2005
`
`Sheet 1 of 3
`
`US 2005/0157675 A1
`
`
`
`
`102
`106
`“Px_y
`
`106A ye
`
`
`
`7 104
`
`AMPLIFIER
`107A
`107 Wey
`
`AMPLIFIER
`
`(9)
`108A
`Vee
`AMPLIFIER
`
`
`
`109 hig 109A |
`
`
`
`AMPLIFIER
`
`
`
`
`
`103
`
`\
`
`\
`te
`
`RADIO TOWER
`
`1114
`
`
`
`105
`
`@
`
`Page 2
`
`CommScope Ex. 1018
`
`

`

`Patent Application Publication
`
`Jul. 21,2005
`
`Sheet 2 of 3
`
`US 2005/0157675 Al
`
`FIC.
`
`2
`
`i:
`
`pail
`\
` 217
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`204
`
`
`
`
`
`
`
`
` )
`205
`
`206
`
`
`
`CommScope Ex. 1018
`
`Page 3
`
`
`
`

`

`Patent Application Publication
`
`Jul. 21,2005
`
`Sheet 3 of 3
`
`US 2005/0157675 Al
`
`- cal 216
`SECTOR a
`
`
`
`
`
`
`
`
`
`
`RRH
`
`215
`
`T 303 ~y
`
`303
`
`
`
`
`218
`
`306
`
`|
`|
`|
`|
`|
`|
`!
`
`mm
`oO
`~~
`
`308
`
`212
`SECTOR’
`
`RRH
`
`Po
`
`ff cy
`304
`
`fs
`
`
`
`
`|
`|
`|
`|
`|
`oof ..
`1312
`310
`
`
`
`
`
`
`
`2
`
`
`
`
`310
`
`
`
`
`
`
`:
`
`
`
`
`
`
`
`[|
`311
`
`
`
`
`
`
`
`3117310
`
`31173
`
`si
`
`|
`|
`|
`|
`|
`|
`|
`|
`|
`|
`|
`!
`|
`|
`|
`|
`_
`
`af]
`
`_
`|
`|
`
`oe kv,
`
`
`
`
`="
`
`it fi
`
`
`
`
`i
`
`a)
`
`3023
`
`rag |}
`
`203
`
`3
`
`FIG.
`
`Page 4
`
`CommScope Ex. 1018
`
`

`

`US 2005/0157675 Al
`
`Jul. 21, 2005
`
`METHOD AND APPARATUS FOR CELLULAR
`COMMUNICATION OVER DATA NETWORKS
`
`CROSS REFERENCE TO RELATED
`APPLICATIONS
`
`This application claims priority to U.S. Provisional
`[0001]
`Patent Application Ser. No. 60/536871 filed Jan. 16, 2004.
`
`FIELD OF THE INVENTION
`
`The present invention is related generally to wire-
`[0002]
`less communications in buildings.
`
`BACKGROUNDS OF THE INVENTION
`
`Wireless communications systems are becoming
`[0003]
`an increasingly integral aspect of modern communications.
`In fact, recent trends show that an increasing number of
`users are replacing all wire-line methods of communications
`with their wireless counterparts such as, for example, cel-
`lular telephones in place of traditional wire-line telephones.
`Since such cellular telephones are essentially radios, it
`is
`well known that signal quality between a cellular base
`station and a handset degrades under certain circumstances.
`The most significant source of degradation occurs when a
`user moves from an outside location to an indoor location
`where the radio signals are required to pass through or
`around various obstructions. Since many users place the
`majority of cellular calls from within buildings or other
`structures, achieving high quality consistent indoor coverage
`is becoming more essential.
`
`Several methods for achieving indoor cellular net-
`[0004]
`work coverage are known. For example, one method of
`achieving such coverage, known as a distributed antenna
`system (DAS), is illustratively shown in FIG. 1.
`A DAS uses
`a base station and a repeater or a power amplifier that is
`typically located within a building to retransmit within the
`building a signal received at an external antenna. Referring
`to the illustrative DAS of FIG. 1, when a signal 103 is
`transmitted from an antenna in a cellular communications
`network, such as antenna 101 (e.g., an antenna in a cellular
`communications network), antenna 113, which is external to
`building 111, receives signal 103. Signal 103 is then passed
`along connection 104 which is, illustratively, a coaxial cable,
`to component 105 which is, in this example, a radio repeater.
`Repeater 105 forwards the signal to amplifiers 106a, 107a,
`108a and 109a. These amplifiers amplify the signal which is
`then transmitted over in-building antennas 106, 107, 108 and
`109. Thus, the result is that cellular telephone 102 receives
`the signal transmitted from antenna 113 via antenna 109. By
`passing the signal along a wired connection from antenna to
`repeater 105 and rebroadcasting the signal over antennas
`106-109, the problems associated with poor signal quality in
`buildings are alleviated.
`
`While DAS systems are advantageous in many
`[0005]
`aspects, they are limited in certain regards. For example, in
`order to install
`a DAS, cabling (such as coaxial cabling)
`must be installed throughout the building at each location
`where an in-building antenna is desired. Thus, installation
`expense is relatively high. Additionally, such systems are not
`flexibly expandable and there is typically no mechanism for
`reprovisioning or reallocating the bandwidth available to
`different locations within the building.
`
`Another method for achieving indoor cellular net-
`[0006]
`work coverage relies on the use of small in-building base
`transceiver stations (BTSs), which are smaller versions of
`well-known base stations such as are used in a traditional
`cellular network, to provide essentially an entire in-building
`cellular network. The result of using such small BTSs is
`a
`network of so-called pico-cells (cells with a short range) that
`operate similarly to a low-powered traditional cellular net-
`work in provisioning bandwidth and managing data and
`voice calls within one or more individual buildings. How-
`ever, since such a system is essentially a miniaturized
`cellular network, management of a multitude of such BTSs
`within a building would be problematic as it would require
`network components (such as a Radio Network Controller
`(RNC) and/or a Mobile Switching Center (MSC) in
`a
`CDMA network) to provision bandwidth and manage calls
`across the large number of pico-cells. Hence, a mini-BTS
`system is relatively cost-prohibitive and complex to install
`and maintain.
`
`a need to
`is
`As cellular usage increases there
`[0007]
`provide increased and cost effective capacity and coverage
`outdoors in dense urban areas, outdoor malls, or in business
`or academic campuses. Many of the same techniques that are
`used indoors can also be used in these environments. Typi-
`cally a base station remotely serves a given outdoor location
`using DAS systems in an architecture known as “hoteling”.
`However, these architectures require the use of proprietary
`RF or fiber links to connect the base stations and the remote
`antennas.
`
`SUMMARY OF THE INVENTION
`
`The aforementioned problems related to in-build-
`[0008]
`ing wireless communications are essentially solved by the
`present invention. In accordance with the principles of the
`present invention, cellular signals or other wireless signals/
`messages are introduced into a building by transmitting
`packetized messages corresponding to those messages over
`a shared or dedicated data network to designated locations
`within the building. Once the designated destination is
`reached, the packet headers are stripped from the packets
`and the wireless message is then broadcast over the air to an
`intended recipient.
`
`first embodiment, base station interface cards
`In a
`[0009]
`(BSIs) are used in place of RF generating equipment in a
`base station such as that used in a cellular communications
`network (e.g.,
`a CDMA network). For downlink signals,
`when a BSI receives coded baseband signals from a pro-
`cessor, such as
`a CDMA Modem Unit (CMU) in
`a CDMA
`system, the BSI then buffers the baseband signals and
`periodically creates data packets each containing a plurality
`of coded baseband signals. The BSI then forwards the data
`packets over a high-speed data network, such as a gigabit
`Ethernet network, to
`one
`or more illustrative
`Gigabit
`switches. These switches duplicate and route the packets to
`one or more specific ports corresponding to
`a cellular
`CDMA sector which in turn corresponds to one or more
`radio transceivers, at least one of which corresponds to the
`address of the intended recipient of the message. In one
`embodiment, the radio transceivers contain equipment that
`extracts the baseband signals from the packets, process the
`base band signal, convert to RF format, amplify the signals,
`and broadcast the signals to
`an intended recipient over
`associated antennas.
`
`Page 5
`
`CommScope Ex. 1018
`
`

`

`US 2005/0157675 Al
`
`Jul. 21, 2005
`
`For uplink signals, the radio transceivers receive
`[0010]
`uplink signals from a mobile user, the transceivers convert
`it
`to
`a digital format and generate packets of the coded
`signals and forward them through the network to the afore-
`mentioned switches and then to
`the BSI and CMU for
`transmission through the traditional wireless network to an
`intended recipient. In another embodiment, lo in order to
`increase the possible number of available radio transceivers,
`one or more special summing nodes sum the base band data
`in incoming uplink packets in order to reduce the number of
`packet streams passing through the BSI to the CMU. Since
`the signals are coded (e.g., with Walsh codes), the CMU can
`differentiate between the signals in the summed data packets
`and forward those signals to an intended destination in the
`wireless network. These summing nodes can be separate
`units, integrated with switches, or their functionality can be
`integrated into one or more of the radio transceivers.
`
`BRIEF DESCRIPTION OF THE DRAWING
`
`FIG.1 shows a prior art distributed antenna system
`
`[0011]
`(DAS);
`FIG. 2 shows an in-building cellular network in
`[0012]
`accordance with the principles of the present invention; and
`
`FIG.3 shows how uplink signals in the network of
`[0013]
`FIG. 2 are aggregated.
`
`DETAILED DESCRIPTION OF THE
`INVENTION
`
`FIG. 2 shows an in-building communications net-
`[0014]
`work in accordance with the principles of
`the present
`invention. In the network of FIG. 2, base station 201 sends
`and receives messages from a wireless network via path 206.
`Base station 201 is, for example, similar to a base station
`used in a traditional CDMA network such as a OneBTS base
`station manufactured by Lucent Technologies, Murray Hill,
`N.J. One skilled in the art will recognize that such a base
`station traditionally has three main components: 1)
`a net-
`work interface for interfacing with the other components of
`the
`cellular network (e.g.,
`a
`radio network controller
`(RNC)); 2)
`a digital baseband shelf typically having a
`processor (e.g.,
`a COMA modem unit (CMU)) for coding
`and decoding incoming and outgoing message traffic, as well
`as
`a radio (e.g.,
`a Universal CDMA Radio (UCR)) for
`modulating/demodulating the coded digital message traffic
`onto/from a carrier signal; and 3) an RF shelf for amplifying
`the modulated signal and transmitting that signal over the air
`to
`a mobile user.
`
`While the base station 201 of FIG. 2 in accordance
`[0015]
`with the principles of the present invention is similar to a
`traditional CDMA base station, there are some differences.
`Specifically, in accordance with the principles of the present
`invention, both the RF shelf and external base station
`antenna functions are not used in the in-building base station
`201. This modified base station 201 may be physically
`located within a building to be serviced by the base station
`or, alternatively, external to that building. Additionally, the
`UCR of the digital baseband shelf 202 in base station 201 is
`replaced by a component herein referred to as a base station
`interface card (BSI) 203. The illustrative BSI 203 may be a
`component of similar form factor to the UCR and illustra-
`tively plugs into the same physical slot in the digital shelf
`traditionally occupied by the UCR. For downlink signals, as
`
`described more fully below, instead of receiving and trans-
`mitting coded cellular radio signals over the air, the BSI 203
`in base station 201 functions to forward mobile user-coded
`baseband signals over a high-speed data network, such as
`illustrative gigabit Ethernet network 208 to a switch/sum-
`ming node, also referred to herein referred to
`as a Radio
`Distributor/Aggregator 210 (RDA), and subsequently to
`a
`desired end destination where the baseband signals are
`converted to RF and then broadcast over the air to a message
`recipient, such as mobile terminal 219. Similarly, as also
`discussed below, on the uplink, when a user, once again such
`as mobile terminal 219,
`transmits messages from, for
`example, a cellular telephone, those messages are transmit-
`ted via the illustrative RDA 210 and illustrative gigabit
`Ethernet network 208 to the BSI 203 and CMU 204 for
`further processing and distribution via the network interface
`equipment 205 and the wireless network via path 206. The
`below illustrative example specifically discusses one imple-
`mentation of the present invention in a CDMA network. One
`skilled in the art, however, will recognize that the principles
`of the invention as herein described will be equally appli-
`cable to
`a GSM, UMTS or other wireless communications
`networks.
`a CDMA system,
`[0016]
`Specifically, on the downlink in
`when a signal is addressed to a mobile terminal in a building,
`such as, referring to FIG. 1, mobile terminal 102 in building
`111, the illustrative BSI 203, discussed above, receives
`digital I and Q baseband signals generated by the CMU 204
`and stores them in a buffer. Illustratively, I and Q signals are
`received every 0.5 microseconds. Once the buffer reaches a
`predetermined level, or a predetermined amount of time has
`passed, the BSI 203 forms an Ethernet packet of those
`signals having a destination address (e.g.,
`a MAC address)
`corresponding to a sector in which an intended recipient is
`present, such as a sector that is distinguished by the time
`offset of the pseudorandom number (PN) code of a pilot
`channel to which the mobile terminal 219 is tuned. One
`skilled in the art will recognize that other configurations and
`operations of the BSI are possible. For example, instead of
`connecting directly to the physical slot in the digital shelf,
`the BSI can be a separate component that connects to
`a
`digital, radio frequency (RF), or intermediate frequency (IF)
`port. In such a configuration, the BSI may receive digital
`coded baseband signals, as described above or, alternatively,
`may receive RF or IF signals and then convert those signals
`to digital form for buffering and packetizing.
`
`The packets are then sent via gigabit Ethernet from
`[0017]
`the BSI to the RDA 210 over, illustratively, gigabit Ethernet.
`For downlink signals, the RDA acts essentially as a switch
`having, illustratively, a plurality of ports. Each port on the
`RDA illustratively corresponds to an addressable sector for
`the routing of messages. For example, each port of an RDA
`may be identified
`as
`a separate sector or,
`if
`a greater
`coverage area is desired, for example, then multiple ports
`may be designated as corresponding to a single sector. Each
`sector, in turn, corresponds to
`a one or more radio trans-
`ceivers, referred to herein as remote radio heads (RRHs),
`such as RRHs 211, 212, 213, 214, 215 and 216 correspond-
`ing to an area of wireless coverage within a building. As one
`skilled in the art will recognize, RDAs can be connected as
`flexibly as regular data switches: multiple RDAs may be
`used in a cascaded fashion to facilitate greater control over
`the routing of messages to end recipients and to permit more
`granularity in the management of bandwidth allocation, or
`
`Page 6
`
`CommScope Ex. 1018
`
`

`

`US 2005/0157675 Al
`
`Jul. 21, 2005
`
`no RDA may be necessary for a point-to-point link between
`the BSI and a particular RRH. Alternatively, in some imple-
`mentations, a single RRH on a single port of the RDA may
`suffice to serve a relatively large sector. One skilled in the art
`will recognize that the number of RRHs necessary to pro-
`vide coverage to a sector will depend upon environmental
`factors such as, illustratively, the number of obstructions
`(e.g., walls or other such obstacles) in proximity to the RRH.
`
`When the RDA 210 receives a message from the
`[0018]
`BSI 203 having a particular address (for example a Medium
`Access Control (MAC) address) corresponding to a particu-
`lar sector, the RDA 210 compares that address to,
`for
`example, a look-up table to identify which ports on the RDA
`210 correspond to the designated sector. This look-up pro-
`cedure can use a variety of existing Ethernet protocols, such
`as using special multicast addresses, or having all RRHs
`belonging to a particular sector be a part of the same virtual
`LAN (VLAN), and broadcasting packets on that VLAN.
`Once the RDA 210 has identified the ports corresponding to
`the recipient sector(s), the
`RDA 210
`will replicate the packet
`(if necessary to forward to multiple end destination RRHs)
`and forward a copy of the packet to the appropriate ports for
`further dissemination to
`the designated sectors and the
`corresponding RRHs 211-216. Each RRH 211-216 has,
`illustratively, network interface equipment, timing and fre-
`quency synchronization equipment, signal processing ele-
`ments, a power amplifier and one or more antennas. The
`network interface equipment of the destination RRH such
`as, in this case, RRH 211 corresponding to mobile user 219,
`receives the packets from the network and removes the
`headers from the packets. The I and Q baseband signals are
`then forwarded to the timing and synchronization equipment
`where the signals are buffered. As described more fully
`below, the signals are then processed, converted to RF
`format and played out to the power amplifier and broadcast
`over the air via the antenna(s) to a recipient end mobile user.
`
`Since CDMA networks and Ethernet networks
`[0019]
`were designed for different uses (i.e., CDMA was designed
`for circuit switched voice applications and Ethernet was
`designed for packet switched data applications), manners of
`transmitting data across those networks differ relative to one
`another. One of the more critical differences is
`in how
`frequency and timing are managed in the different networks.
`Specifically, CDMA networks were designed with a tight
`timing/jitter tolerance of less than 2-3 microseconds using a
`synchronous frequency as required by the air-interface.
`Ethernet, on the other hand, was designed with a loose
`timing/jitter tolerance and an asynchronous frequency that is
`adequate
`for packet-switched data networking in,
`for
`example, a star network configuration. Overcoming these
`timing and frequency differences to achieve synchronization
`is critical to passing timely packets of CDMA data over an
`Ethernet network.
`
`More specifically, timing synchronization is espe-
`[0020]
`cially crucial for downlink traffic since the offsets in the pilot
`channels are used to identify the base station sectors in the
`network. On the uplink, however, while timing synchroni-
`zation is important, one skilled in the art will recognize that
`it
`is sufficient to assure a certain, fixed delay among the
`uplink signals from the RRHs 211-216—precisely synchro-
`nizing the exact time is not necessary. In both uplink and
`downlink scenarios, timing synchronization may illustra-
`tively be achieved by first determining the minimum feasible
`
`time that a packet will spend transiting the data network
`between the BSI 203 and the RRH 210, hereinafter referred
`to as the minimum packet delay, T,,,;,. This minimum packet
`delay is, for example, measured as a function of delays in
`buffering baseband signals in
`the BSI 203 to form the
`packets, transmission through the gigabit Ethernet MAC, the
`physical layer and the switches in the RDA 210, as well as
`the delay experienced traveling over cables. Thus, t,,i, 1S a
`basic, illustrative reference time between the BSI 203 and
`the RRH 210 for a given Ethernet network topology.
`
`This minimum reference time t,,;,, of course, is not
`[0021]
`the time typically experienced by a packet transiting the data
`network, only the feasible minimum based on known delays.
`The actual timing delays of packets through the data net-
`work are
`a function of,
`in part, queuing delays in the
`presence of other data traffic in the data network. This actual
`timing delay may vary from one packet to the next and, over
`a given number of packets, a spread in the timing delay, At,
`can be determined. Over a sufficient number of packets, the
`spread At can be lo measured such that a maximum timing
`delay of T,,:,,AT may be determined.
`‘min+’
`
`Therefore, in order to broadcast packets from an
`[0022]
`RRH, such as RRH 210, to mobile 219, and ensure conti-
`nuity between the packets transiting the network, a timing
`delay can illustratively be introduced into the RRH broad-
`cast such that each successive packet is guaranteed to be
`present at the RRH and ready for broadcast at its appointed
`time. Specifically, if the RRH broadcast delay is established
`at a time greater than T,,;,,At, with, illustratively, an addi-
`tional timing tolerance delay added to
`t,,;,,
`to manage
`additional timing jitter, then each packet will be at the RRH
`when it is scheduled for broadcast.
`
`The timing delay and spread can be determined by
`[0023]
`either a hardware or a software solution, or a combination of
`both. In an exemplary hardware solution, the timing delay
`can be calculated through knowledge of the topology of the
`network and the delay properties of the routing equipment
`and cables. By setting higher priority to
`CDMA packets, and
`knowing maximum allowable packet length (for example,
`1500 bytes), one can predict the maximum delay spread At.
`Alternatively, timing delays and spread can be directly
`measured using a variety of software methods, such as
`methods involving an exchange of time stamps between
`endpoints and using statistical techniques to determine the
`time delay and spread. One illustrative example of such a
`software method is the well known Network Time Protocol
`(NTP). One skilled in the art will recognize that many such
`hardware and software methods may be used to determine
`the timing delay and spread.
`
`As previously mentioned, in addition to timing
`[0024]
`synchronization, frequency synchronization across the net-
`work is
`also
`important. Frequency synchronization is
`achieved in accordance with the principles of the present
`invention by either a hardware solution or a software solu-
`tion. An illustrative hardware solution to frequency synchro-
`nization is achieved by using the physical layer in Gigabit
`Ethernet networks to synchronize the CDMA signals. In
`order to achieve this synchronization, a frequency oscillator
`is illustratively used in each RRH as a frequency reference
`for all the frequency synthesizers in the RRH. These fre-
`quency oscillators are locked, using well known clock and
`data recovery methods, to the clock rate of the data coming
`
`Page 7
`
`CommScope Ex. 1018
`
`

`

`US 2005/0157675 Al
`
`Jul. 21, 2005
`
`in on the Ethernet connection. This clock rate, in turn, is set
`by the clock which BSI 203 uses to encode the Ethernet
`signals. The BSI 203 can use, for example, a stable and
`accurate stand-alone reference oscillator to generate all its
`clocks or, alternatively, may derive its reference from the
`base station clock. One skilled in the art will recognize that
`such a hardware implementation of frequency synchroniza-
`tion will require forwarding a frequency reference from each
`network node, thus requiring additional overhead to main-
`tain synchronization in this manner. In addition, one skilled
`in the art will recognize that similar synchronization tech-
`niques may be used in other network transport methods.
`
`On the other hand, if a software solution to fre-
`[0025]
`quency synchronization is used, a timestamp is illustratively
`applied by the BSI 203 to each downlink CDMA packet
`marking the time it is transmitted from BSI 203. At the RRH
`210, the arrival time of each CDMA packet is recorded using
`a local clock and the difference between the embedded time
`stamp and the measured arrival time is calculated as the
`delay D. If the illustrative clocks at the BSI and RRH are
`synchronized, then D should be held constant. Thus the local
`RRH clock can be adjusted to the remote BSI 203 clock
`using well-known statistical methods. For example, a num-
`ber of timestamp-minus-local-clock measurements corre-
`sponding to multiple packets can be used to calculate the
`frequency deviation over time. Over a desired period of
`time, the frequency error can be inferred from the total delay
`change that is detected. A frequency correction correspond-
`ing
`to
`this frequency error is used compensate for
`the
`frequency deviation. In this way, the software controls the
`RRH 210 local clock by tracking and correcting the fre-
`quency in relation to the BSI 203 clock. Furthermore, to
`reduce the packet arrival jitter, and thus the accuracy of the
`frequency tracking mechanism, only specific packets can be
`used for frequency tracking. More particularly, in
`this
`example, the delay is recorded over a selected number of
`packets and is used for software synchronization. By selec-
`tively using an lo ensemble of packets with a measured
`delay, such as the smallest delay, the effect of switch jitters
`due to background traffic may be significantly reduced.
`Uplink signals are transmitted in
`a similar fashion as
`described above in association with downlink signals. When
`an uplink signal is received by an RRH, such as RRH 211,
`from, for example, illustrative mobile terminal 219, that
`RRH is will convert the signal to a digital format and will
`buffer the digital signals and packetize them at
`a predeter-
`mined time interval or until a predetermined buffer fill level
`reached. The RRH 211 will then send the packets of digital
`signals to
`the corresponding RDA 210 to which it
`is
`attached. As discussed above, each RDA 210 can have a
`plurality of ports associated with a plurality of addressable
`sectors. And, as also discussed above, multiple RDAs can be
`cascaded to in a way such that numerous RDA ports can be
`addressed to even more numerous remote radio heads.
`However, the CMU 204 typically can only accept a smaller,
`limited number of signal channels (e.g., 6 channels). This is
`not an issue on the downlink as the RDAs simply replicate
`the downlink packets and retransmit identical packets to
`multiple addresses. However, on the uplink, the packets
`flowing into the RDAs to the BSI 203 and CMU 204 are not
`identical—they are potentially each from different mobile
`users. Thus, a problem arises as
`to how to reduce the
`
`potential relatively large number of unique uplink packet
`data streams into the limited number of channels acceptable
`to the CMU 204.
`
`This problem is overcome by the principles of the
`[0026]
`present invention. Specifically, referring to FIG. 3, in order
`for a large number of RDAs to communicate with the CMU
`in BSI 203, an illustrative processor at each RDA in
`a
`cascaded structure of RDAs, such as RDA 210, strips the
`headers off the packets corresponding to
`a unique sector,
`sums the data in those packets, and repacketizes the data
`before forwarding it to another RDA or the BSI 203. One
`skilled in the art will recognize that this processor may not
`be located within the RDA and that it may also be located in
`an independent unit attached to the RDA. Alternatively, the
`functionality of this processor may be located within one or
`more of the RRHs. In the case of cascaded RDAs each
`having such a processor, each RDA in the cascade will
`accomplish this summing function until the packets flowing
`to from a particular RDA 210 to the BSI 203 on the uplink
`correspond only to one of the channels of the CMU. While
`summing the data in packets traditionally would result in
`irreparably destroying the packets (i.e., because reconstruc-
`tion of the original packets would be impossible), the data in
`the packets of the present invention correspond to coded
`baseband signals of all the traffic on a particular sector. Thus,
`using traditional processing well known in the art, the CMU
`can process and identify signals in the summed packets
`corresponding to
`a uniquely CDMA coded mobile user.
`These unique signals are then forwarded to end destinations
`within the cellular network.
`
`More particularly, FIG. 3 shows an in-building
`[0027]
`network having two illustrative sectors, 307 and 308, each
`comprising two RRHs, each of which supports two traffic
`carriers (frequencies). As described above, on the downlink
`data packets 301-304 are transmitted by BSI 203 in direction
`305 to RDA 210. The RDA then uses the addresses on each
`of those packets to determine to which sector (and in turn to
`which RRHs) a packet is routed. As shown in FIG. 3,
`packets 302 and 304 are, illustratively, encode information
`for traffic carriers in sector 307. Accordingly, RDA 210
`replicates packets 302 and 304 and multicasts those packets
`in direction 306 to both RRH 212 and RRH 218 for
`radiation. Similarly, packets 301 and 303 are multicast in
`direction 309 to RRH 215 and RRH 216 which in turn will
`radiate these traffic carriers in sector 308.
`
`On the uplink, however, and also as briefly dis-
`[0028]
`cussed above, the RDA 210 acts as a summing node and
`aggregates the uplink data packets. Specifically, when RRHs
`support several mobile carriers (frequencies), those RRHs
`will each buffer the packets and forward them to RDA 210.
`Thus, for example, RRHs 218 and 212 will send out packets
`310 and 311 in direction 312 to RDA 210 and RRH 215 will
`illustratively send packets 314 and 315 in direction 313 to
`RDA 210. When those packets arrive at
`the RDA, the
`headers of the packets are removed and the data in any
`packets in the same sector are summed together. The RDA
`210 then interleaves all unique baseband signals together
`and forwards out a single data stream containing packets of
`baseband signals from all sectors of the RDA 210. In
`a
`cascade of RDAs, this process is repeated at each RDA so
`that only a desired number of data streams containing
`packets of baseband symbols arrive at the BSI 203 in FIG.
`2 and the CMU 204 in FIG. 2
`in the base station. One skilled
`
`Page 8
`
`CommScope Ex. 1018
`
`

`

`US 2005/0157675 Al
`
`Jul. 21, 2005
`
`in the art will recognize that such a summing function is
`illustrative and may not be necessary depending upon the
`particular implementation of the network described herein.
`Additionally, if such a summing function is used, one skilled
`in the art will realize that this function can be performed at
`network nodes other than the RDA, such as at an RRH.
`
`One skilled in the art will recognize that the above
`[0029]
`net