Transmit Noise Floor

The fundamental lower bound of the transmit noise floor is thermal noise, defined by the equation kTB (Boltzmann's constant × temperature × bandwidth). At room temperature (290K), this equates to -174 dBm/Hz. Every active component in the transmitter chain—amplifiers, mixers, and oscillators—adds its own noise figure (NF), elevating the noise floor above this theoretical minimum. The cumulative noise figure of the transmit lineup directly determines the broadband noise power delivered to the antenna.

Local oscillator (LO) phase noise is a dominant contributor to the transmit noise floor, particularly at frequency offsets close to the carrier. Imperfections in the LO signal are directly transferred to the upconverted RF output, creating a noise pedestal around the transmitted carrier. This mechanism is distinct from nonlinear spectral regrowth. Even a perfectly linear amplifier will exhibit an elevated noise floor if driven by a noisy LO. Close-in phase noise can desensitize receivers operating in immediately adjacent channels.

It is critical to distinguish the transmit noise floor from spectral regrowth. The noise floor is present even with no modulated signal applied—it is the residual broadband noise of the active circuitry. Spectral regrowth, conversely, is signal-dependent distortion caused by power amplifier nonlinearity that broadens the occupied bandwidth of a modulated carrier. While both degrade Adjacent Channel Leakage Ratio (ACLR), they require different mitigation strategies: linearization for regrowth, and low-noise design with filtering for the noise floor.

The transmit noise floor becomes a critical problem when it leaks into a co-located receiver's band, a phenomenon known as receiver desensitization or self-blocking. Even if the transmitter is operating on a different frequency, its broadband noise can overwhelm a sensitive receiver attempting to detect a weak signal. The required isolation between transmitter and receiver is calculated as: P_noise_tx - (Sensitivity + SNR_min). This dictates duplexer and filter specifications in Frequency Division Duplex (FDD) systems.

In modern direct-conversion transmitters, the digital-to-analog converter (DAC) sets a practical noise floor limit. The DAC's quantization noise and jitter-induced noise create a broadband noise pedestal that propagates through the analog chain. The signal-to-quantization-noise ratio (SQNR) for an ideal N-bit DAC is approximately 6.02N + 1.76 dB. Oversampling and noise shaping techniques, such as delta-sigma modulation, can push this quantization noise out of the band of interest, lowering the effective in-band noise floor.

The primary defense against an excessive transmit noise floor is high-selectivity filtering after the power amplifier. In FDD systems, a duplexer provides frequency-dependent isolation between the transmit and receive paths. The duplexer's stopband attenuation in the receive band must be sufficient to suppress the amplified transmit noise floor below the receiver's sensitivity threshold. This requirement often dictates the use of high-Q technologies such as Surface Acoustic Wave (SAW) or Bulk Acoustic Wave (BAW) filters.

Unwanted radio frequency energy generated by a transmitter at frequencies outside the occupied bandwidth and adjacent channels. Unlike ACLR, which focuses on immediate adjacent channel leakage, SEM limits apply to a much wider frequency range, including harmonic frequencies. Transmit noise floor is a primary contributor to SEM, and regulatory bodies like the FCC and ETSI define strict SEM masks that must be met. These limits protect distant spectrum users, such as public safety or radio astronomy services, from interference caused by broadband noise and spurious tones.

The distribution of a signal's power as a function of frequency, measured in dBm/Hz. PSD provides the fundamental visualization for assessing the transmit noise floor. A clean transmitter will show a sharp drop in PSD outside the occupied bandwidth, limited only by the thermal noise floor and phase noise. When nonlinearities are present, spectral regrowth elevates the PSD in adjacent channels. Engineers use a spectrum analyzer to measure PSD and verify that the noise floor meets the required spectral mask.

A critical metric quantifying the ratio of transmitted power within an assigned channel to the power leaking into adjacent channels. A poor transmit noise floor directly degrades ACLR. It is the primary regulatory compliance measure for spectral regrowth. ACLR is typically measured for the first adjacent channel (ACLR1) and the second adjacent channel (ACLR2). Modern 5G NR specifications demand ACLR values exceeding 45 dB, requiring highly linear power amplifiers and effective digital pre-distortion (DPD) to suppress the noise floor.

A power amplifier phenomenon where the current output depends on past input states due to thermal, electrical, or trapping dynamics. This causes frequency-dependent nonlinear behavior that complicates the cancellation of spectral regrowth. Memory effects cause an asymmetry in the spectral regrowth sidebands, meaning the transmit noise floor may be higher on one side of the carrier than the other. Compensating for these effects requires advanced DPD models like the Volterra series or memory polynomial, which can predict and cancel the distortion dynamically.

Nonlinear amplitude-to-phase conversion where the phase shift introduced by a power amplifier varies with the instantaneous input signal envelope. This is a critical source of spectral asymmetry and regrowth. While AM-AM distortion causes gain compression, AM-PM distortion causes a phase shift that depends on signal power, generating intermodulation products that directly raise the transmit noise floor. Modern GaN-based amplifiers exhibit significant AM-PM, making its accurate modeling and correction essential for meeting strict spectral mask requirements.

A signal processing technique that intentionally redistributes quantization or clipping noise energy away from critical in-band frequencies to less sensitive out-of-band regions. In the context of the transmit noise floor, noise shaping is often used in conjunction with Crest Factor Reduction (CFR). When a signal is clipped to reduce PAPR, the resulting distortion noise is shaped via filtering to fall outside of the critical adjacent channels, effectively lowering the ACLR without degrading in-band Error Vector Magnitude (EVM).