In a multi constellation context, demanding phases of flight such as APV operations can be targeted using Receiver Autonomous Integrity Monitoring (RAIM) to check GNSS integrity. This is why it is necessary to characterize all failure modes inducing range errors of the order of a few meters. One of these failure modes is radio frequency interference which causes very penalizing errors since they can affect several measurements at the same time. This work focuses on advanced and classical GPS/Galileo RAIM performance degradation evaluation in presence of multiple failures due to interference when stringent phases of flight are targeted. The way the User Equivalent Range Error UERE, especially the receiver noise residual error, is computed is first reminded in this paper. The main unintentional interference sources to be accounted for in the ARNS are then given and their impact is discussed. More particularly an expression of the code tracking error envelope in presence of CW interference is proposed. A complete pseudo range measurements model is given taking into account the interference effect. The issue of integrity monitoring in presence of interference is studied considering two different algorithms: the Sequential Constrained Generalized Likelihood Ratio Test based RAIM and the Snapshot Least Squares Residuals RAIM. The last part of this paper is dedicated to GPS/Galileo RAIM simulations that have been conducted using this proposed model on GPS L1/L5 and Galileo E1/E5b pseudorange measurements under different conditions: nominal ones, with a CW interference with a power within the GPS L1 C/A interference mask or 20dB above this mask. CW were added on two frequencies: the worst Galileo spectrum line and the worst GPS spectrum line. The two integrity monitoring algorithms formerly described have been tested for two modes of flight APV 1 and APV2. The goal of these simulations is not to detect interference, it is to observe RAIM capacity to detect a bias, corresponding to a satellite failure with a probability of occurrence of 10-4 /h, in presence of interference It is seen that there is no impact of CW interference within the interference mask. Both algorithms studied manage to detect dangerous biases with the required probability and their actual performance are not degraded. On the contrary, CW interference with a power 20 dB above the mask has an important effect on pseudo range error variance and can even lead to misleading situation using classical snapshot RAIM. However, in any case, if advanced methods of monitoring integrity are used that is to say measuring the ability of detecting dangerous biases, the user will be protected from hazardous misleading situation.