The original Data Over Cable Service Interface Specification (DOCSIS®) has been revised a number of times since its approval as an international standard in 1998. After the first revision, the original DOCSIS became known as version 1.0 and the succeeding version numbers are 1.1, 2.0, 3.0, and 3.1. Each revision was intended to address advancements made by the cable industry’s competitors and to accommodate the demands of the marketplace. A feature of all DOCSIS versions is that the hardware operates with previous versions. This allows an operator to support all earlier versions of DOCSIS in their network. For example, a DOCSIS 3.1 modem works as a DOCSIS 2.0 modem if 2.0 is the current version running on a network. Supporting all versions also allows operators to make a gradual transition to the newer DOCSIS version instead of doing a massive replacement of modems. Because of the likelihood that different versions of DOCSIS modems are being used, installers and technicians should be able to recognize the features and capabilities of each version.

Introducing DOCSIS Modems

Until 1998, Internet access using a dial-up telephone modem was the only practical method available to residential customers. However, many people became acquainted with higher levels of Internet access while at work because businesses typically subscribed to integrated services digital network (ISDN) and T-1 service connections for high-speed data (HSD) access. Thus, there was already an increased demand for HSD service by the time cable operators introduced Data Over Cable Service Interface Specification (DOCSIS) cable modems to their customers.

DOCSIS was developed with the intention that it could operate on legacy tree-and-branch and hybrid fiber/coax (HFC) architectures. In both architectures, the downstream DOCSIS carrier is assigned to a 6 megahertz (MHz) wide quadrature amplitude modulation (QAM) channel in the range from 88 to 860 MHz. Because the DOCSIS channel is shared by a number of subscribers, when each modem is activated and assigned to a customer account, the modem is configured so that it recognizes only those data packets addressed to it. All other packets are rejected. To ensure that all customers can access the shared channel, the maximum downstream throughput for each modem is configured at a rate that is substantially lower than the throughput of one 64- or 256-QAM channel (Table 1). These throughput rates are configured by the operator to be much higher than the maximum rate of dial-up modems.

Upstream data is sent from the cable modem to a cable modem termination system (CMTS) that is located in the headend using the radio frequency (RF) return spectrum between 5 and 42 MHz. The RF return spectrum is vulnerable to interference and noise funneling from customer premises; both of these can prevent a signal from being received in the headend. Because using this portion of the spectrum required vigilant maintenance, before DOCSIS, the RF return spectrum was seldom used by many cable operators. In the upstream, DOCSIS utilizes modulation schemes and narrower bandwidths that are likely to be received despite the presence of interference and noise in the network. Quadrature phase shift keying (QPSK and 16-QAM are the modulation schemes used to address these requirements, which are specified with bandwidths ranging from 0.2 to 3.2 MHz. The maximum data throughput of a 16-QAM carrier with a bandwidth of 3.2 MHz is 10.24 Mbps. Under most circumstances, customers receive much more data than they send (Figure 1). Therefore, the unequal, or asymmetrical maximum downstream and upstream data rates are normally acceptable.

Figure 1: Asymmetrical downstream and upstream data rates.

It is impractical and unnecessary for the modems to be continuously sending data upstream. Instead, a communications access technology called time division multiple access (TDMA) is used to divide the upstream frequency spectrum into timeslots. Each modem is then given a timeslot to transmit its data upstream. In this manner, several hundred modems can share the same upstream channel by using it at different times.

Enhancing the Capabilities of DOCSIS

Cable operators aspired to compete with local telephone companies long before DOCSIS modems become a reality. With the introduction of DOCSIS modems, voice over Internet protocol (VoIP) became a technological alternative to how the telephone companies provided service. All that the customer needs is a subscription to a VoIP provider and a multimedia terminal adapter (MTA) to convert their voice to IP data packets when speaking, and to convert IP data packets to voice when listening.

By the time DOCSIS modems became available, VoIP technology already existed, and several companies were  offering inexpensive VoIP telephone service. The Internet; however, is a best-effort transport medium, which means that telephone data packets can be delayed or lost, causing telephone calls to become distorted or even disconnected. These VoIP telephone competitors offered low prices as a way to lure customers away from the traditional telephone companies, but with compromised quality and reliability.

Cable operators chose to compete by offering a high-quality and reliable product that rivals all of the products of the traditional telephone companies at a competitive price. To do this, CableLabs created PacketCable, an IP-based platform for telephone service over a DOCSIS access network. A requirement of PacketCable called for quality of service (QoS) controls (beyond the best-effort delivery of data), enhanced security, and adaptive equalization in the upstream path for improved data throughput. Consequently, a sub-revision of the original DOCSIS standard, called DOCSIS 1.1, was released in 1999. Table 1 shows what major changes were made to the original DOCSIS standard (version 1.0). Because DOCSIS 1.1 is a sub-revision of the original DOCSIS standard, many discussions use DOCSIS 1.x in reference to both standards.