Support


USR6000 Series Cable Modem Support
This page has been designed for printability. To print this document, click the Print button on your browser's toolbar. When you are finished, click the Close Window button located at the upper-right or bottom of the page to close this window and return to the previous page.

DHCP   TOD   SNMP   Web   Email   News Protocols on top of DOCSIS
UDP   TCP
Internet Protocol (IP)
Data Link Encryption (BPI)DOCSIS specific protocols
Transmission Convergence (MPEG-2)
(downstream)
Media Access Control (MAC)
(upstream)
Physical Layer Modulation (PHY)
DOCSIS Protocols

DOCSIS Protocol Stack

The various protocols that are used in the DOCSIS standard all have specific functions. The relationships of these protocols are shown in the diagram labeled "DOCSIS Protocols". These protocols are the basis of operation for any cable modem. The lower four layers are specific to the cable data network (HFC plant) and are the foundation of communication between any cable modem (CM) and the cable modem termination system (CMTS). The upper layers are protocols that are carried over the communication path established by the lower layers. Well known IP services such as email, web, file transfer, and network news are all presented as TCP/IP traffic. Other protocols such as IPX/SPX, AppleTalk, and NetBeui are possible however since a vast majority of the internet is based on TCP/IP, they are not shown here.

The lower four layers of DOCSIS are:

PHY - physical layer. This layer defines the modulation schemes used on a HFC plant. This layer is responsible for translating the raw signal found on the actual coaxial cable to and from digital information used by the upper layers in the protocol stack. The modulation schemes present on the coaxial cable can be 64 or 256 quadrature amplitude modulation (64- or 256-QAM) for the downstream and quadrature phase shift keying or 16-QAM for the upstream.

MPEG-2 - transmission convergence layer. All data that is present on the downstream is encapsulated into MPEG-2 frames. These frames can contain the actual video and audio data that is typically decoded and presented as TV image and sound. They can also contain data that is decoded and presented as information available for computer usage (i.e. the internet). Since all data... voice, audio, internet... is encoded into MPEG-2 frames, it is quite feasible for a cable operator to multiplex all signals onto one cable. This is inherently how cable internet access works.

MAC - media access control layer. All data that is present on the upstream is managed by this section of the DOCSIS protocol stack. Since a HFC plant is similar to an Ethernet network in that all communication devices are "connected" to the same cable, it is imperative that an orderly process exist whereby the CMTS can tell the CM when to transmit data and for how long. The MAC layer is the means of coordinating upstream traffic from the CM to the CMTS.

BPI - data link encryption. Since the cable network is a shared medium, there must be a method to protect user data from malicious usage. The DOCSIS standard defines Baseline Privacy Interface as this method.

Lock, Range, Register, Forward

From the moment a CM is powered on to the time the end-user is able to load their favorite home page, a sequence of events occurs that involves all layers of the DOCSIS protocol stack. They are very specifically designed to ensure that the cable operator can provide a stable and reliable connection to the internet. This document seeks to explain the interaction of these layers to provide for a stable data connection between the CM and the CMTS. A detailed discussion of the physical, mac, mpeg-2, and mac protocols is beyond the scope of this document. For the purpose of this explanation, enough information is provided for a basic understanding of these protocols.

QAM Lock

Assuming that the CM has properly booted and passes all of its internal diagnostics, the first thing a CM must do is find a "DOCSIS channel".

Practically every HFC plant in North America is based on a 6 Mhz channel plan. If you tune your TV to channel 32, the cable box is actually presenting a 6 Mhz slice of the total frequency spectrum. That 6 Mhz is used to encode MPEG-2 frames containing both video and audio information that your cable box or TV decodes into picture and sound. If you graphed a single channel provided by the cable operator, it would look similar to this...

For each channel, repeat that graph and place them end to end, left to right, lowest channel to highest channel. That is the frequency spectrum for the cable operator. The far edge of that graph to the left is 5 Mhz. The far edge of that graph to the right extends beyond 1000 MHz however it is uncommon for a cable operator to operate any channels beyond 900 Mhz.

A DOCSIS channel can be graphed in the same fashion. Instead of video and audio information inside the MPEG-2 frames, it contains data that represents computer information. Every MPEG-2 frame is stamped with a program identifier (PID). Any MPEG-2 frame with a PID of 0x1ffe (hex) is a DOCSIS data frame.

A CM will scan the entire frequency spectrum looking for a 6 MHz channel where the MPEG-2 frames are stamped with a PID of 0x1ffe. Once the CM finds a 6 Mhz channel that carries this type of MPEG-2 frame, it has achieved what is called QAM lock. Once QAM Lock is obtained, a CM proceeds to the next step in the process.

Ranging

Every CMTS continuously broadcasts three separate messages on every downstream that relate to the MAC layer. These are SYNC, UCD, and MAP. A CM must continuously identify and decode these messages in order to properly transmit data onto the cable upstream path.

The SYNC message is simple a time synchronization message. It is used to provide a common time reference for all cable modems connected to the CMTS via the coaxial cable.

The next message is the upstream channel descriptor message (UCD). It is essentially a description of the upstream parameters including modulation, symbol rate, channel width, and frequency. Each upstream that is available on the cable plant has its own set of these characteristics. For each upstream, a separate UCD message is broadcast by the CMTS. It is imperative that the CM use this to match the available upstream characteristics to its own capabilities.

The final message is the bandwidth allocation map (MAP). Like the UCD message, every upstream has its own MAP message. The MAP contains information that indicates when a CM can transmit and for how long. The SYNC message provides the time reference for these transmission opportunities. For the very first transmission that a CM makes, it must reference the MAP for a transmit opportunity called the initial maintenance opportunity - this is specifically reserved for CMs that are just connecting to the network after power up.

Once a CM has received and decoded all of these messages it "knows" the following:

  • Time reference

  • Upstream parameters

  • When to transmit

With this information, the CM proceeds to the next step of the ranging process which is summarized as:

  • Fine adjustment of the time reference

  • Fine adjustment of the transmit frequency

  • Fine adjustment of the transmit power

The process of ranging is iterative. The CM will use the SYNC, UCD, and MAP messages to determine basically when and how it transmits information on the upstream. The CMTS will see the CM transmission, determine the time offset from when it sent the MAP to the CM, the exact frequency of the incoming transmission, and the power level of the incoming transmission. The CMTS will then send a ranging response to the CM with instructions to adjust the timing, frequency, and power level of the CM transmission. This exchange repeats until the CMTS and CM have properly adjusted the timing to within 1µs, frequency within 10 Hz, and power within .25 db.

During the length of time that a CM is actively connected to a CMTS, the CMTS will periodically repeat this process. This is called periodic maintenance and is determined by the configuration of the CMTS.

Once ranging is complete, any further transmission by the CM onto the upstream is negotiated based on a request from the CM that is granted by the CMTS. The CM will request bandwidth during a transmit opportunity as defined in the MAP message. The CMTS will grant the request and schedule a unique opportunity for the CM to transmit. This process takes place on a time scale µs. As you can imagine, coordination of time and opportunity is critical. This is why the SYNC message is so critical to upstream communication.

Register

Now that the cable modem has established the lower layers of communication with the CMTS, its next step is to establish IP connectivity. This is done through standard DHCP procedures. The CM broadcasts a DHCP request. The CMTS will forward this request to a DHCP server. The DHCP server will, in the most basic of systems, register the unique CM by its Ethernet MAC address (different than the MAC layer in the DOCSIS protocol model) and assign it an IP address. This is the DHCP response which is forwarded by the CMTS. This response contains at least the IP address, default gateway, and IP net mask.

In addition to this basic IP information, the DHCP response will also contain DOCSIS specific information necessary for basic CM operation:

  • TFTP server ip address

  • DOCSIS configuration file name

  • time-of-day server (optional)

  • cable operator network specific information (optional)

The TFTP server is located on the cable operator's network and is where the various DOCSIS configuration files are maintained. These configuration files specify operating parameters for the CM. The most basic configuration file simply enables the CM and declares a maximum available bandwidth for both the downstream and upstream. In more complex cable systems, these configuration files can be customized for specific vendor's modems, certain classes of service, special operating instructions such as traffic filters and restrictions, and to initiate a firmware upgrade of the CM.

To download the configuration file, the DHCP response sent to the CM contains the TFTP server ip address and configuration file name to be downloaded. Once the CM configures itself with the IP address provided by the DHCP server, it then initiates a TFTP file transfer to the TFTP server and downloads the configuration file specified. Once the CM has downloaded the configuration file, it is decoded and processed by the CM.

Data Ready!

Now the CM is ready for data forwarding mode. Most cable modems designed today employ a bridge architecture. In the most simplest terms, a bridge device is simply one that converts from one media to another. For a CM, the two medias are, on the CPE side - Ethernet or USB, on the RF side - coax. Any defined data traffic that is presented on side of the cable modem is simply forwarded, or bridged, to the other side.

Was this FAQ helpful? (1 = not helpful, 5 = very helpful)
1   2   3   4   5  
 
Did it resolve your problem?
Yes Yes   No   Not sure (haven't tried it yet)  
 
Additional Comments (255 characters maximum)

Close Window