An HF beacon network has been deployed in eastern Alaska to estimate plasma number densities in the ionospheric volume overhead. The network includes three dual-frequency CW transmitters employing pseudorandom noise (PRN) codes and three receivers. The network spans latitudes 62–68° north geographic. Observables from the network include the signal amplitude, Doppler shift, and group delay of all the ray paths joining the transmitters and receivers with a 1 min Cadence. The data are interpreted using a forward model based on geometric optics and an inverse method which solves for the parametrized ionospheric plasma number density in the volume overhead as an optimization problem. Sensitivity analysis is incorporated through automatic differentiation. Additional information from regional GNSS receivers, ionosondes, and incoherent scatter radars can be included in the inverse method. Range and Doppler spreading of the HF signals can furthermore be interpreted in terms of ionospheric turbulence and irregularities along the ray paths. This paper describes the methodology involved and compares electron density estimates with measurements from the Poker Flat Incoherent Scatter Radar (PFISR) which were withheld from the inverse methodology for validation.