In this paper, an integrated sensing and communication (ISAC) system was proposed under cell-free architecture with different power allocation schemes for both communication and sensing functions. In addition to using uplink (UL) pilot transmission for channel estimation, a group of pre-designed orthogonal downlink (DL) pilot sequences are inserted and transmitted from each cooperative access point (AP) towards the user equipments (UEs). The inserted DL pilots in cell-free ISAC are utilized for both channel estimation in communication and cooperative localization in radar sensing. Based on the ISAC strategy, a power allocation scheme is proposed with the objective of minimizing the maximum channel estimation error for each user, under the power constraints of APs and minimum mean square error constraints of the localization. The full-power scheme which APs consume all the available power during the DL phase is also studied and analyzed. The performance of the proposed ISAC strategy is evaluated under a realistic highway scenario, and compared to the existing non-DL strategies. Additionally, we investigate the impact of the pilot ratio under a constrained signal length, as well as the impact of the pilot proportion that the pilot length to the entire signal frame length. Simulation results show that the performances of both channel estimation and localization can be significantly improved at a relatively low cost of extra power needed for APs. Moreover, the analysis provides insights into optimizing the DL/UL pilot ratio and pilot proportion which can significantly enhance both channel estimation and localization accuracy, while ensuring sufficient resources remain available for data payload.

In this paper, downlink (DL) pilot based integrated sensing and communication (ISAC) is explored in cell-free networks. To improve both accuracy of channel estimation and localization in the ISAC based cell free network, a group of pre-designed DL orthogonal pilot sequences are introduced in the proposed ISAC strategy for both channel estimation for communications and cooperative sensing with radars. A power allocation scheme is proposed with the objective of minimizing the maximum channel estimation error for each user, under the power constraints of APs and minimum mean square error of sensing based localization. The performance of the proposed ISAC strategy is evaluated under a realistic highway scenario, and compared to the conventional methods. The results show that the performances of both channel estimation and localization can be significantly improved at a relatively low cost of extra power needed for APs near users.