3.3 Physical Layer

Role of the Physical Layer

The Physical layer is the first and lowest layer of the OSI model. Its job is to take bits that upper layers want to send and actually move them across a physical medium, such as copper wire, fiber optic cable, or the air in wireless networks. Everything at this layer is about electrical, optical, or radio signals, not about addresses, frames, or packets. Those logical structures belong to higher layers.

At this layer, the network is only concerned with how a 0 or a 1 is represented and transmitted, and how the receiving side can correctly detect and interpret those bits. It deals with the mechanics of transmission, not with meaning or routing.

Bits, Signals, and Media

Although the upper layers think in terms of bits, the physical world only has signals. The Physical layer defines how to map bits to signals and back again. A bit is an abstract concept. A signal is something that can be measured, such as voltage on a wire, light intensity in a fiber, or radio wave strength and phase in wireless.

For each physical medium, the Physical layer specifies how a binary 0 and a binary 1 look. On copper, that might be one voltage level for 0 and a different level for 1. On fiber, that might be light off for 0 and light on for 1, or more complex changes in wavelength or phase. On wireless, it might be different changes in frequency, amplitude, or phase of a radio carrier signal.

Different media are suited to different distances, speeds, and environments. Copper cables are common in local networks, fiber optics excel at high speed and long distance, and wireless is convenient for mobility and areas without cabling, but each of these choices still performs the same essential job at this layer, which is to carry signals representing bits.

Encoding and Clocking

If two devices are going to exchange bits reliably, they must agree on exactly how bits are encoded onto the medium and when each bit starts and ends. This is where line encoding and timing, also called clocking, come in.

Line encoding is the scheme that turns a stream of 0s and 1s into a pattern of physical signal changes. A very simple idea would be to use a constant high level for 1 and a constant low level for 0, but this creates problems for long runs of identical bits, which can make it difficult for the receiver to keep track of timing. More sophisticated schemes deliberately introduce changes in the signal, even for repeated bits, so that the receiver can always find bit boundaries and keep its clock synchronized.

Clocking refers to the rate at which bits are sent, for example, 1 Gbit per second, and how the sender and receiver stay in sync. In some systems, a separate clock signal is provided. In many modern systems, the clock is embedded inside the data signal, and the receiver recovers timing information from the pattern of changes in the signal. If the clocks drift and lose synchronization, the receiver will start sampling at the wrong moments and bits will be misread.

Encoding and clocking are invisible to higher layers. As long as the Physical layer delivers the correct sequence of bits, upper layers do not need to know how the bits were represented on the wire or how the timing was maintained.

Bit Rate and Signaling Rate

The speed of a physical link can be described in two related but distinct ways, bit rate and signaling rate. The bit rate is how many bits per second are transmitted. The signaling rate, often measured in baud, is how many signal changes, or symbols, occur per second on the medium.

If each symbol only carries 1 bit of information, the bit rate equals the baud rate. Modern systems often encode multiple bits in a single symbol, for example 2 bits per symbol or even more. In that case, the bit rate is higher than the symbol rate.

A simple relationship holds when each symbol carries the same number of bits. If we call the number of bits per symbol $k$, the symbol rate in baud $S$, and the bit rate in bits per second $R_b$, then:

By increasing the number of bits per symbol, a system can increase its bit rate without increasing the number of signal changes per second. This can improve efficiency but often requires cleaner channels and more complex hardware.

Physical Layer Devices

Some network devices operate purely at the Physical layer and do not understand frames, MAC addresses, or IP addresses. They only care about the signal itself and how to pass it along or regenerate it.

Classic hubs in early Ethernet networks are an example. A hub receives incoming electrical signals on one port, regenerates and amplifies them, and repeats them out of all other ports. It does not inspect or filter frames and does not make any decisions based on addressing. From the hub’s perspective, everything is just a stream of bits.

Repeaters are another Physical layer device. When signals travel over a medium, they weaken and degrade. A repeater receives these weakened signals, reconstructs the original signal shape as closely as possible, and transmits a refreshed version. This allows longer distances than a single cable segment would support before signal quality becomes too poor.

Media converters also belong at this layer. They take a signal in one physical form, such as electrical signals on copper, and convert it to another, such as light in a fiber. The content of the data is not interpreted. The converter only changes the physical representation of the bits.

Because Physical layer devices never look inside frames or packets, they preserve all higher layer information unchanged. Their role is to maintain signal quality, extend distance, or bridge different media types.

Physical Layer Specifications

The Physical layer is defined by a collection of characteristics that together describe exactly how data moves between devices. These characteristics form part of standards that manufacturers use to build compatible hardware.

Voltage levels, or corresponding parameters for optical and radio systems, are defined so that transmitters and receivers from different vendors can interoperate. The standard specifies what range of values counts as a valid 0 or 1 at the receiver. For copper, this might be a specific nominal voltage with tolerances. For optical systems, it might be a minimum and maximum optical power level for detection.

Connectors and pinouts are also part of the Physical layer. For example, a common copper Ethernet connector has a specific shape and size and a defined order of wires on its pins. Matching connectors ensure that cables and ports can be physically joined and that signals reach the correct conductors.

The maximum cable length for reliable transmission is specified so that designers know how far devices can be separated without excessive signal loss or distortion. If the length limit is exceeded, the receiver may see too much noise or too weak a signal, which leads to bit errors.

Other characteristics can include the type of encoding to use, the signaling rate, electrical impedance of cables, and how devices must behave at link startup. All of these details are kept at the Physical layer so that upper layers can remain general and independent of the medium.

Simple Full-Duplex Concept

Although the concept of duplex modes is explored separately, the Physical layer is where the actual capability to send and receive at the same time is realized. In a simple half duplex link, the Physical layer only allows transmission in one direction at a time along the same path. In a full duplex link, the Physical layer is designed so that the two ends can send and receive simultaneously, either by using separate physical paths in each direction or by using signal processing techniques to separate the two directions on the same medium.

The rules for when a station can transmit, how it detects if the path is busy, and how simultaneous transmissions are avoided belong higher up, but the fundamental ability is provided here. If the Physical layer is only capable of one direction at a time, no higher layer logic can turn that into true full duplex.

Interaction with the Data Link Layer

The interface between the Physical layer and the Data Link layer is very simple. The Data Link layer hands a sequence of bits to the Physical layer for transmission. The Physical layer turns those bits into an appropriate series of signals on the medium. On the receiving side, the Physical layer observes the medium, recovers the clock, detects the symbols, converts them to bits, and passes the resulting bit stream up to the Data Link layer.

The Physical layer does not know about frames or addresses, even though the bit sequence it is transmitting or receiving contains frame headers and checks. From its point of view, all bits are equal. Any decisions about what a sequence of bits means, or whether it should be forwarded or discarded, are deferred to higher layers.

Because of this clear division, changes in one physical technology, for example from copper to fiber, do not require changes to Ethernet frame structure or higher level protocols. As long as the Physical layer continues to deliver the same bit sequences faithfully, upper layers can function without modification.