Having previously discussed aspects of multiplexing in this column, this month we’ll consider the technology in broader terms. In general, multiplexing is the process of transmitting multiple signals within a common channel, with the purpose of using a shared medium. On the receiving end, a demultiplexer (demux) extracts the individual signals that were originally combined in a
multiplexer (mux).
Multiplexing can increase channel utilization by multiple programs or optimize channel use by a single program, or both. The amount of bandwidth needed to transmit a multiplex must be higher than that of the constituent signals (usually a straight sum of the individual bandwidths, plus any additional overhead), although the increase may be offset by means of information-reducing techniques, such as compression.
So the usual goal of multiplexing is not to save bandwidth, but to use it efficiently and to share a common, typically scarce or expensive, medium. It is also possible to split a single source signal into multiple lower-rate signals, transmit those and then perform the inverse function at the receiving end — a process called Inverse Multiplexing (IMUX).
Multiplexing can be divided into three general types: time domain, frequency domain and spatial domain. Time Division Multiplexing (TDM) is essentially the time-based switching of multiple analog or digital input signals. In the digital realm, we can extend the utility of TDM to that of switched information packets within larger streams. With the ubiquitous MPEG standards, packets of information (each potentially carrying different content) are multiplexed within a program stream or transport stream, and statistical multiplexers (stat mux) can dynamically adjust the amount of bandwidth needed by each constituent program within a larger stream, making for efficient use of a limited-bandwidth “pipe.”
Other forms of content distribution use multiplexing, too. Within storage devices or information streams, multimedia files can combine various video, audio and data components into one package or file format, forming a pre-multiplexed container. Ethernet is another TDM means for carrying multiple packets from and to different destinations using a common cable, often with routers (another form of multiplexer) along the way. And, pixel-based displays use multiplexing order to minimize the number of interconnections to the display-driving transistors.
Frequency Division Multiplexing (FDM) is widely used in broadcasting and networking. FDM involves the use of multiple carriers to transmit multiple signals over a common medium, usually cable or wireless spectrum. This involves the process of modulation, whereby the information signal is impressed onto the carrier signal.
One common form of FDM is Orthogonal FDM (OFDM), used for DVB-T, DSL modems and various 802.11-based wireless local area networks. Another variation of FDM is Wavelength Division Multiplexing (WDM), which places information onto multiple optical carriers within a beam of light carried in free space or over an optical fiber.
Spatial Division Multiplexing (SDM) is a technique whereby multiple RF signals occupy the same bandwidth (or carrier frequency), but in a different physical space. Cross-polarization is one long-established variety, whereby independent signals are transmitted on the same carrier frequency by using orthogonal electric field vectors, commonly referred to as horizontal and vertical. When received by orthogonal antennas, the two signals can be separately detected and demodulated. In addition to placing two signals on orthogonal planes, the carriers are often displaced by 90 degrees of electrical phase shift as well, resulting in Circular Polarization (CP), which has been shown to alleviate some of the effects of multipath interference.
Another form of SDM exploits a characteristic called Orbital Angular Momentum (OAM), which operates by placing separate signals on multiple elements of a specially constructed helical antenna. Sometimes called “twisted radio beams,” OAM emerged from research in optical communications and is different from CP. With CP, the electric field vector rotates in space, whereas with OAM, the wave front rotates. OAM signals have been generated by using antenna arrays consisting of concentric uniform circular arrays, with the antenna elements in the arrays fed with a successive delay from element to element. Although OAM is believed by some to offer substantial gains in bandwidth efficiency, others remain skeptical, describing it as a subset of Multiple-Input-Multiple-Output (MIMO) at best, with equivalent properties and no added advantage.
New multiplexing technologies are also emerging. MIMO is an SDM technology that increases transmission capacity by using spatially separated transmission antennas together with spatially separated receiving antennas. Currently used in 802.11n wireless technology, MIMO works by splitting a source signal into several lower-rate components and transmitting those components in the same frequency channel from spatially separated antennas, as shown in Figure 1. The transmitters radiate into the broadcast area on the same frequency, and multiple antennas are used to receive the transmitted signals, which arrive from different directions. (Note that the input/output terminology refers to the “air” medium, and not the source or destination equipment.)
MIMO technology is based on an earlier concept called Distributed-Transmit-Directional-Receive (DTDR). In some respects, the Single-frequency Network (SFN) and distributed transmission modes of digital terrestrial television share similarities with MIMO, the key difference being that multiple receive antennas are not used with digital transmission systems such as DVB and ATSC.
MIMO is not diversity reception, which uses multiple receiving antennas to combat multipath, but only one transmit antenna. Nor is it a directional array, which uses multiple transmitting antennas to generate a directional transmission pattern. And is not a combination of both.
One intriguing aspect of MIMO technology is the ability to operate in a closed-loop fashion, i.e., with feedback. When one-to-one communications are used (such as for unicast and peer-to-peer links) and a return channel is available, receivers can send, back to the source, channel information regarding scattering, fading and power decay with distance. This information can then be used by an intelligent transmitter to change coding parameters that can maximize the use of the channel, and minimize the received error rate.
Combined with large-area self-configuring networks, emerging multiplexing technologies are hoped to offer significant gains in bandwidth efficiency — a national resource that is in ever-increasing demand.
—Aldo Cugnini is a consultant in the digital television industry.