To reduce the cryogenic cooling cost for high operating temperature applications, the dark current of infrared photodetectors need to be further reduced. Many methods have been proposed such as inserting unipolar-barrier, complementary barriers, the M-structure, and double heterostructures. The Interband Cascade Infrared Photodetectors (ICIP), originally arising from the interband cascade laser, has also been tried. The ICIP structure consists of high bandgap AlSb material, which can further reduce the dark current. Therefore, these features may make high temperature operation possible for ICIPs, especially for the mid wavelength range.This work investigates the ICIP for mid wavelength operation. The ICIP is designed as a p-i-n type. Specifically, the i region is composed of a 5-stage interband cascade structure. Each cascade stage consists of an electron-barrier (eB) layer, an absorber layer, and a hole-barrier (hB) layer, and this sequence repeats five times. The photogenerated electrons in one absorber layer should first pass through the hB layer by optical phonon assisted stepwise transport. The carriers, which arrive at the hole states of the eB layer, finally tunnel to the next absorber layer. Consequently, the energy levels of the eB and the hB layer should be equidistributed. In other words, the energy difference between adjacent levels should be comparable to the optical phonon energy, about 30 meV.The absorber layer of each period is designed as 0.5 μm thick InAs (2.4 nm)/GaSb (3.6 nm) SLs, which has an expected cut-off wavelength of around 4.28 μm. The hB layer is made up of 8 InAs QWs, separated by AlSb barrier layers. For the eB layer, it consists of AlSb(2.1 nm)/GaSb(5.3 nm)/AlSb(2.1 nm)/GaSb(7.5 nm)/AlSb(2.1 nm). The calculated energy separation between adjacent levels is close to the optical phonon energy. The sample is grown on an n-type GaSb (001) substrate by molecular beam epitaxy. After the growth, an array mesa was formed using standard photolithography and dry-etched.We measured the temperature-dependent dark current to reveal the dark current mechanism of the device. The dark current of the device is dominated by the diffusion current instead of the generation-recombination current for the temperature range between 180 K and 300 K, as seen from the Arrhenius plot. The spectral responsivity is measured by calibrating the photocurrent spectrum with the blackbody response with the blackbody source temperature set at 800 K. We calculated the detectivity of the device. At 77 K, the 50% cutoff wavelength is 4.02 μm and the detectivity D* is 1.26×10¹² cm·Hz^1/2/W for the peak wavelength of 3.79 μm at 0 V. At 300 K, the 50% cutoff wavelength is 4.88 μm and the D* is 1.28×10⁹ cm·Hz^1/2/W for the peak response wavelength of 4.47 μm at 0 V.We also observed the Negative Difference Resistance (NDR) effect in our device. The NDR effect can be seen in the dark current curves from 77 K to 220 K. The I_P (the NDR peak point current) has almost no change while the I_V (the NDR valley point current) increases, and therefore the Peak-to-Valley Current Ratio (PVCR) (I_P/I_V) decreases with increasing temperature. There are two parts of the tunneling process arising from the interband cascade structures: tunneling through the electron barriers and resonant tunneling. As the dark current equation, $n(V)$ has a strong influence on the dark current, which is determined by the product of the density of states $N\left(E\right),$ and the Fermi-Dirac distribution function $f\left(E\right)$, and the tunneling probability $T\left(E,V\right)$. At the valley point, the dark current is mainly caused by the barrier tunneling mechanism. For the barrier tunneling current, the product of the density of states and the Fermi-Dirac distribution function $N\left(E\right)f\left(E\right)$ increases with elevating the temperature. The tunneling probability, $T\left(E,V\right),$is exponentially proportional to the energy state. That is to say, the $T\left(E,V\right)$ at a higher energy state is larger than that at a lower energy state. As a result, the I_V rises. Meanwhile, the $T\left(E,V\right)$ of the resonant tunneling dark current can be treated as a constant and the $N\left(E\right)f\left(E\right)$ decreases, resulting in a drop of resonant tunneling current. The dark current at the NDR peak point (I_P) is caused by the both two tunneling mechanism, leading to an almost constant peak dark current. Correspondingly, an unchanged I_P and an increased I_V lead to a smaller PVCR for a higher temperature.