How the Cable TV is Used for Internet Access
By Gabriel Torres on September 10, 2013
The cable TV network can be used for connecting a computer or a local network to the Internet, competing directly with DSL (Digital Subscriber Line) technology. In this tutorial, we will teach you how this kind of connection works.
This type of network is classified as HFC (Hybrid Fiber-Coaxial), as it uses both fiber optics and coaxial cables. The connection between the cable TV company to the distribution points (also called optical nodes) is made using fiber optics, with distances up to 25 miles (40 km), while the connection between optical nodes and the users’ home is made using coaxial cables. Each optical node typically serves between 500 and 2,000 clients. In the coaxial network, amplifiers can be used in order to regenerate the signal and expand the maximum length of the coaxial network. Because of this configuration, cable TV networks do not offer problems with the quality of the cable, the length of the cable, or electromagnetic interferences (EMI) that plague typical DSL systems, in particular ADLS systems, as they use conventional telephone cables.
The most common system used by cable TV companies to offer Internet access is called DOCSIS (Data Over Cable Service Interface Specification). Currently, there are four DOCSIS versions available: 1.0, 1.1., 2.0, and 3.0. Versions 1.0 and 1.1 are identical, except that version 1.1 accepts quality of service (QoS) parameters, allowing the use of the cable TV network for telephony (voice over IP, a.k.a. VoIP). Therefore, we call versions 1.0 and 1.1 simply 1.x. Version 3.0 allows channel bonding, as we will explain later on.
The coaxial cable used by the cable TV allows broadband communication, i.e., the transmission of several channels using distinct frequencies. Typically, each channel is 6 MHz wide, and a whole channel is used for downstream, i.e., download, with a maximum theoretical transfer rate of 42.88 Mbps (38 Mbps available to the user) in all DOCSIS versions. Of course, the transfer rate you will reach depends on the service you hired. For downstream, DOCSIS uses quadrature amplitude modulation (QAM), using either 64 or 256 levels.
For upstream, i.e., upload, the configuration used depends on the DOCSIS version. Usually, a channel between 200 kHz and 3.2 MHz wide (depending on the speed hired by the user) is used in version 1.x, allowing a maximum theoretical transfer rate of 10.24 Mbps (9 Mbps available to the user). DOCSIS 2.0 and 3.0 use a 6.4 MHz channel, which increases the maximum theoretical transfer rate to 30.72 Mbps (27 Mbps available to the user). Keep in mind that the speed you actually achieve depends on the speed you hired from the cable TV company. Upstream signals use quadrature phase-shift keying (QPSK) or quadrature amplitude modulation (QAM), depending on the speed.
In theory, the service does not need to be asymmetrical (i.e., different upload and download speeds). For instance, it is possible for a cable TV company to offer a service with 5 Mbps for both upstream and downstream, even using DOCSIS 1.x.
DOCSIS 1.x uses time-division multiple access (TDMA) technique, while DOCSIS 2.0 and 3.0 the code-division multiple access (CDMA) technique is also available.
The version 3.0 of DOCSIS allows the use of more than one channel at the same time, feature known as channel bonding. This drastically increases the transfer rates that can be achieved: you can simply multiply the maximum theoretical transfer rate by the number of channels used. For example, if four channels are used for downstream, the maximum theoretical transfer rate is increased to 171.52 Mbps. Once again, the actual speed you will achieve depends on the cable TV company. It is interesting to note that DOCSIS 3.0 does not set a limit of channels that can be bonded together; however, the cable TV company is limited to the equipment available on the market.
The actual transfer rate achieved using cable TV networks is related to the number of users connected to the optical node at the same time, as this system is based on the fact that not all users will be accessing the Internet at the same time. More users mean lower bandwidth available to each individual user.