5G RACH preamble detection is crucial for the following 5gnetwork KPIs: Random Access Success Rate (RASC) as well as Accessibility Success Rate (ASR) and Session Setup Success Rate (SSSR).
RACH preamble detection follows special rules and procedures which are not 3GPP standardized but rather left to the vendor’s baseband unit chip-set.
A thorough explanation to the underlying RACH preamble detection in the overall RACH procedure is the key aspect to understand 5G RACH False Detection, ghost detection and RACH misdetections. This would potentially lead to successfully optimized 5G networks optimization.
Intoduction: 5G RACH preamble detection is crucial for the following 5G network KPIs: Random Access Success Rate (RASC) as well as Accessibility Success Rate (ASR) and Session Setup Success Rate (SSSR).
RACH preamble detection follows special rules and procedures which are not #3GPP standardized but rather left to the vendor’s baseband unit chip-set.
A thorough explanation to the underlying RACH preamble detection in the overall RACH procedure is the key aspect to understand 5G RACH False Detection, ghost detection and RACH misdetections. This would potentially lead to successfully optimized 5G networks Optimization.
Overview: Random Access is an important aspect of mobile systems like LTE and 5G, where multiple users are always competing in a CBRA technique for resources. However, background noise (i.e. noise due to high temperature) and interference caused by own cell (i.e. multiple users trying to access on the same RACH occasion using different preamble sequences), neighbor cells (i.e. cells on the same frequency with different RACH Root sequence planning, or overshooting cells on the same frequency with same RACH Root Sequence planning) or other networks and sources, imposes a significant problem to those systems causing them to falsely detect access requests. Consequently unnecessary processing leading to processor overload and air traffic signaling overload are generated based upon these unreal request events.
RACH detection problem: In the Physical Random Access Channel (PRACH), cyclic-shifted versions of Zadoff-Chu (ZC) sequences are adopted as preambles, based on the fact that ZC sequences present Constant Amplitude and Zero Auto-Correlation (CAZAC) properties. A Constant Amplitude Zero Auto-Correlation waveform (CAZAC) is a periodic complex-valued signal with out-of-phase (cyclic-shift) periodic autocorrelations equal to zero. Even when different root sequences are used, the cross-correlation exhibits high discrimination ability among the available preambles. These properties of CAZAC Zadoff-Chu sequences turn them ideal to generate different preambles based on cyclic-shifted versions of the same root ZC sequence.
A 5G/LTE UE can generate preambles by randomly selecting different root ZC sequences or by applying random cyclic shifts to the same ZC sequence, as part of the general RACH Root sequence planning principles.
gNB should be equipped with a specific Random Access signaling processing unit detector. This is important since when detecting multiple random access requests from users in a mobile network, the task of such a Detector is to decide whether only noise floor or RACH preamble requests-plus noise and interference are present. In general Interference from other cells is scrambled in 5G and LTE with the different PCI planning and would appear as uncorrelated noise, but not always depending on the transmitted power.
For multiple RACH preamble attempts in a cell, based on traffic load, a RACH detector typically should be able to determine if multiple accesses to the network are been requested based upon statistical computations of the cross-correlation function between the hypothetical received PRACH preamble sequence signal on the proper RACH occasion and a root ZC sequence. The results from this statistical analysis are employed to calculate a threshold value that is then used to decide whether there are users requesting access or not. Correct selection of the threshold is very important once it determines the probability of false alarm, as well as the probability of detection. The correct determination of a threshold value that produces both low false alarm rate and good detection rate is then of extreme importance, under the uncertainty of the noise and multiple RACH preamble attempts interference variance.
Detailed analysis: The implementation of the Random Access processing unit detector depends on the vendor (i.e. baseband unit Altio Star proprietary implementation) and neither GSMA nor 3GPP have any suggestion on that circuitry implementation. However the basic modules, based on typical signal processing, should employ:
– A down-converter to shift the received PRACH preamble pass-band signal to base-band signal for further filtering.
– A linear filter to avoid aliasing after decimation.
– A CP removal module fed with the result of the previous decimation block
– FFT module to transform the SC-FDMA symbols from time domain into frequency domain.
– A sub-carrier de-mapping block extracting the RACH preamble sequence from the output of the FFT module.
– The signal generated by the prior sub-carrier de-mapping is multiplied (cross-correlation operation) by a root ZC sequence. A Zero-padding module fed by the multiplication result.
– An IFFT module transforming the cross-correlation result from frequency domain into time domain. After those signaling processing steps, all samples coming out of the final IFFT block have their square modulus (square root of the sum of the squares of the real part and the imaginary part of the complex number) calculated producing the Power Delay Profile (PDP) samples.
– A preamble detection block to estimate and jointly determine the channel propagation delay (ToA) in the presence of noise, interference and carrier frequency offset Δf due to channel impairments and circuitry PLL offsets failures. This jointly estimated J(ToA, Δf) is the threshold for RACH preamble detection.
As a result then, detection block compares the sample’s PDP against the jointly detected ToA and Δf and decides whether a preamble is present or not. If a preamble is positively detected then as an output of the detection process this detection block sends to the MAC layer the Signatures, Cv and Channel Propagation Delay ToA estimates of all detected preambles.
Detection impairments: Out of this detection procedure two general error events may arise with this threshold-based preamble detection:
– RACH preamble detection failure (Misdetection): The random access preamble is present, but the statistical jointly estimation J(ToA,∆f) does not exceed the detection threshold.
o The discrimination threshold is based on the vendor’s circuitry implementation. There is not much to do on such event, rather than setting the RACH preamble parameter to proper values for individual channels. (A mathematical study and analysis might be a good point to start)
– False alarm: There is not any real random access preamble present in a specific RACH occasion, however statistical jointly estimation J(ToA,∆f) exceeds the detection threshold due to noise/interference variations. Follows some comments:
o Typically the more overshooters and the more intra-cell RACH preamble attempts are present per RACH occasion, the more is the probability to false alarm detection (a mathematical model might reveal this relationship)
o Clearly, there exists a trade-off in setting the detection threshold but it is again vendor proprietary imolementation. Increasing the detection threshold lowers the false alarm rate at the cost of increased likelihood of misdetection.
o For random access preamble detection in 5G cellular systems, the detection threshold is usually chosen such that the false alarm rate is below some target, risking then the misdetection condition (vendor proprietary selection in baseband unit processor).