WO2002063855A2 - Synchronized half-duplex, residential tv coax, digital high-speed multimedia local area communication system - Google Patents

Abstract

The present invention makes use of shared media, such as a residential passive coax distribution in which isolation is purposely provided between user devices, such as TV sets or VCRs, to prevent echoes or other cross-interference. This new approach overcomes the isolation between end user ports by providing a hub or central device at the entry point of services. This will allow an optimum cost-effective solution for high-speed data distribution in a shared media environment.

Description

SYNCHRONIZED HALF-DUPLEX, RESIDENTIAL TV COAX,

DIGITAL HIGH-SPEED MULTIMEDIA

LOCAL AREA COMMUNICATION SYSTEM

The present application claims priority to previous provisional patent application, dated February 05, 2001 and transmitted by US Express Mail using Mailing label # EL043165293US. The present application represents a continuation of the above referenced patent application.

Overview

The proposed system makes use of shared media, such as a residential passive coax distribution in which isolation is purposely provided between user devices, such as TV sets or VCRs, to prevent echoes or other cross-interference. This new approach overcomes the isolation between end user ports, defined hereinafter as Client Data Interfaces (GDI), by providing a hub or central device, hereinafter defined as Hub/Gateway (HG), at the entry point of services, such as CATN, as illustrated in Figure 1.

The system uses a single spectrum that is time-shared for upstream and downstream data traffic, also control and status information.

The system uses frequency-synchronous half-duplex burst transmission to provide two- way transmission within the passive coax distribution system. Specifically downstream bursts are used to carry data or other information (in packets) between the HG to multiple CDIs and downstream request bursts are used to initiate carriage of data or other information (in bursts) from the CDIs to the HG. Data that is intended to pass between the CDIs is received by the HG, checked for errors and then forwarded downstream to the CDIs. Address information contained in the downstream information ensures that only the required GDI (or CDIs) receive and pass on data to the PC or other User Device, hereinafter defined as UD.

Connectivity between the HG and external networks, such as the Internet, may be provided, for example, by one or more Ethernet-equipped interfaces - such as a cable Modem, DSL Modem, wireless or fiber.

Synchronous Burst Operation

Downstream and upstream traffic between the HG and CDIs occurs whenever there is traffic or control information to be transacted. An example of the formats and timing for downstream frequency-synchronous burst transmission is illustrated in figure 2. In this example, the downstream Frame (HG Transmission) contains 80 Bytes and comprises: a SYNC byte containing a fixed and unscrambled value, such as 47π_e ; a TYPE/FLAG byte which identifies the purpose of the packet, e.g. packet Data, CBR Data or control - the FLAG portion of this byte may be used, for example, to indicate if the previous upstream and/or downstream burst contained errors; an address (ADR) used to route the data or control information to a specific device (or group of devices in the case of broadcast or narrowcast data) The next part of the Downstream Frame comprises: a 64 Byte data field and a two-byte check sequence. The downstream transmission is then terminated to allow time for devices to acknowledge correct receipt of their information.

To facilitate constant bit rate (CBR) transmission of Telephony PCM samples the frame frequency should be an integer multiple of 125 microseconds, the international standard period between PCM speech samples.

The transmission and reception of bursts of information between the HG and CDIs can be simplified by locking the clock frequency and phase of the CDIs to that of the regular downstream burst clock frequency. Use of low-cost quartz crystal or ceramic oscillators with digitally stored frequency control or correction would be preferred solutions.

Transmission Formats and Timing

Downstream Data

In format A shown in Figure 2, packets of data or control information, transmitted to a specific CDI and received delayed due to finite propagation delay of the coax wiring (format B), are checked by that addressed CDI and, if the address plus data agree with the check- word, a short acknowledgement burst is returned (format C) to the HG. If the check- word does not correspond with that calculated from the downstream address plus data, or if the downstream address is 00κe or invalid, no acknowledgement burst is returned.

To allow for group addressing of multiple CDIs (data broadcasting) a field in the TYPE byte indicates this condition and inhibits the transmission of a ACK burst.

After propagation delay the acknowledgement (ACK) burst arrives at the HG (format D). If no ACK were received the HG would normally set the TYPE/FLAG byte to indicate that the next burst is a re-send of the previous information or data to the CDI.

Upstream Data Request and Response

In format A shown in Figure 3, a short sequence requesting upstream data be returned is transmitted to a specific CDI and received delayed due to finite propagation delay of the coax wiring. All CDIs detect the TYPE+FLAG byte and prepare to send upstream data or status information. However, only that CDI which detects its own address and performs a successful check will transmit a response. The upstream response comprises a preamble to prepare the HG to receive correctly; the CDI address (which must match that contained in the request burst); a data field (in this example, 64 bytes in length) and finally a check sequence. At this point the CDI will stop transmission in order that the HG can take its turn to transmit.

In the event that no response is received, or that the check word does not correspond with the address+data, the HG would normally set the TYPE+FLAG byte to indicate a re- request, such that the addressed CDI would repeat the last data transaction.

PCM Telephony Transaction

In format A shown in Figure 4, the TYPE+FLAG byte indicates that the first half of the transmission frame be used for downstream PCM samples and the second half of this frame be allocated for corresponding upstream PCM samples. This is followed by a sequence of pairs of CDI Address and PCM sample bytes. In this example, up to 12 telephone channels may be carried downstream and upstream within a single frame.

Frames of this type would normally occur every 125 microseconds, thus meeting the internationally agreed telephony speech-sampling rate. If more than 12 telephone channels are required, additional frames containing different CDI addresses may be inserted. Other frames, not used for telephony, may carry data, control or status information.

CDIs, upon detecting a telephony transaction frame, check the following (alternate) Address bytes. Upon detecting its own address, a CDI accepts the next downstream PCM sample and prepares to send an upstream PCM sample.

The CDIs which receive downstream PCM samples transmit their upstream PCM samples in an order and timing such that they are received successively, and with the correct timing, at the HG. To achieve this, the CDIs, during address allocation, receive commands from the HG, which establish a specific delay (stored in a register in the CDI) between receipt and sending of PCM samples.

Each upstream PCM sample is preceded with a short pre-amble that is used to "fine-tune" the timing of the clock used to receive the PCM sample.

CDI Control / Status Inquiry

The format used for telephony as illustrated in Figure 4. may also be used to set controls in, and receive status information from up to 12 CDIs while using only a single frame period, thus reducing the control/status loading on the transmission system and thus improving traffic utilization of the spectrum. To indicate this mode, a different TYPE+FLAG byte is used.

Packet Segmentation and Re-assembly

The data capacity of each burst is 64 Bytes, thus corresponding to the size of the shortest Ethernet packet. Ethernet (or PPP) packets of greater length are segmented, transmitted and received 64 bytes at a time and then re-assembled before being passed on by this proposed system.

Data "Bridging"

Bursts flowing from a CDI to an external network are re-assembled within the gateway function of the HG to form standard Ethernet or PPP packets. The HG maintains a table of IP addresses of UDs attached to UDIs. Bursts containing the start of an Ethernet (or PPP) packet are identified in order that the IP address may be extracted and the entire packet, contained in one or more bursts, routed to the destination CDI.

Packets transmitted between CDIs on the same passive coax may be passed, burst-by- burst, via the HG. Buffering and packet flow management is provided within the HG.

Packet Transmission and Flow Control

In the event that a CDI buffer feeding a CI nears overflow, the ACK byte may be changed to another value, thus indicating to the HG to pause transmission to that CDI.

Upstream Traffic Flow

CDIs are "polled" to ascertain whether they have data waiting to be sent upstream. Those having data can then be specifically requested to return a data burst. Even upstream traffic performance across CDIs may be achieved by serving those having data waiting, in sequential order, prior to performing another "poll". By polling those CDIs which have just sent data more frequently than the remaining CDIs, emphasis may be given to the CDIs having the most upstream data traffic, thus minimizing latency.

It is possible to provide upstream traffic management by over-riding the above upstream polling sequence. Similarly, the burst rate of downstream packets may be throttled to provide specific traffic limitations.

Traffic Management

For most applications, means should be provided to manage average and/or peak traffic flow rates for each CDI in each traffic direction. Such traffic flow control may be imposed by associating two count memory locations with each CDI, one for upstream, the other for downstream.

The values of these counters may be incremented on a regular basis, at a frequency corresponding to the mean traffic rate required, and decremented each time a 64-byte data packet is transmitted. Thus, for example, if count incrementation takes place every 500 micro-seconds, the mean traffic flow will be limited to 128 K Bytes/second. If the offered traffic rate causes the count to be decremented faster than its rate of incrementation, the count will reach zero. While the count remains at zero, the traffic loses its turn to be transmitted and has to wait. If the count becomes incremented (to value = 1) before the next time that the data has its turn to be transmitted, the next 64-byte data packet will be transmitted, otherwise it again loses its turn. In this way the traffic becomes regulated.

The period of averaging for bursty data is dependant directly upon the count numeric range. Thus a maximum count value control becomes a means of controlling the traffic- averaging period.

Providing two or more count locations associated with each CDI traffic direction may be used to impose multiple traffic flow criteria. While any of these count locations remains at zero, traffic loses its turn to be transmitted. Thus different count increment rates and maximum values may be used to control traffic in a variety of ways.

Control of peak traffic rate is another option that may be achieved using the above method. In this case the maximum count value would be = 1 and the rate of incrementing the count would provide a direct control on the peak traffic rate.

Carriage of Constant Bit-rate Traffic

To provide for CDIs carrying upstream constant bit-rate (CBR) traffic, such as streaming video, audio or telephony, these CDIs will always be polled for such traffic only (defined in TYPE), at regular intervals according to capacity and/or latency traffic specifications. This polling may take priority over that of the regular upstream data polling and traffic management. Thus CBR traffic will use that capacity it requires and the remaining system capacity becomes available for packet data.

Error Containment

In the event that an upstream burst from a CDI contains errors in data or address fields, that CDI will be immediately re-polled and receive a request to re-transmit its previously transmitted burst. This action takes place independently of any traffic flow controls. The actual number of re-transmissions requested may be limited, for example to perhaps four times.

Similarly, any downstream packets, other than those containing telephony or, for example, downstream video streaming (as defined in the TYPE+FLAG bytes) may be retransmitted if no acknowledgement occurs due to errors in either address or data.

System Spectrum and Capacity

In passive coax distribution systems that also carry TN channels, in some cases up to a frequency of 860 MHz, it is convenient to use spectrum above 900 MHz for digital multimedia communication. Coax splitters and directional taps operate sufficiently well, for digital transmission, at frequencies up to 1 GHz. Thus the single spectrum in this proposed system would conveniently be contained between 900 MHz and 1 GHz. Even with simple modulation sehemes, total capacities in excess of 50 Mbit/s may be realized with available and economic technologies. Thus this proposed system may be utilized for multiple purposes including high-quality digital video and audio, high-speed computer packet communication, as well as lower speed, latency sensitive traffic, such as telephony.

Transmission scrambling is employed to ensure good spectral shaping and to ensure sufficient bit timing information at the receivers. A simple and economical method of scrambling is to create and store in HG memory, a table of scramble code sequence based on a well-known or standard algorithm a d to exclusively OR (XOR) the downstream frame, bit-by-bit as it is transmitted. An identical table is stored in the CDI is used to XOR the incoming signal following the SYNC byte, thus restoring the bytes to their original transmitted states. Similarly, the CDI uses its scramble code table to scramble the upsteam burst following its preamble and the HG uses its scramble code table to unscramble the upstream bursts from a CDI. In the example given in this document he pfe-amble signal starts with a 1010 binary sequence and ends in 1100 to denote that the next byte contains the scrambled transmitting CDI address.

Claims

What is Claimed is:

1. A system that provides digital communication between user devices in a coax distribution system engineered to create isolation between analog user entertainment devices, such as TVs, through use of a hub device that receives, buffers and routes data between data user devices.

2. The synchronous and time-shared transmission of data in a shared, passive, distribution network using a single frequency spectrum. The method employed uses a fixed frame length, request and response approach, for upstream data transmission.

3. A combination method of providing average traffic flow, multiple average traffic flow and peak traffic flow management.

4. This invention will allow for different bit rates and modulation methods between the upstream and downstream directions of transmission, preferably such that there is a locked bit-rate relationship between the two directions of transmission.

5. This invention will allow for different transmission symbol rates between the upstream and downstream directions of transmission, providing a common frame timing relationship between the downstream and upstream transmissions is maintained, thus allowing the interleaving of data and control/flow information..

6. This invention will allow for the the carriage of any data format or protocol beyond that of Ethernet, PPP or individual or blocks of PCM speech samples. A further example being the carriage of MPEG-2 formatted digital video bursts.

7. This invention will allow for the carriage of a plurality of different data protocols. The protocol used may be identified, by way of example, either by the use of distinct burst TYPE codes or by specific ranges of address codes. Thus the system may simultaneously used to carry, for example, Ethernet, USB, serial or parallel PPP, Firewire, Appletalk, Asynchronous Transfer Mode (ATM), MPEG-2 digital video and telephony.

8. The invention as described in the specification and referenced figures.