The rapid expansion of Internet of Things (IoT) applications has led to significant access challenges for radio access networks, potentially causing bottlenecks and degrading service quality. This paper proposes Dual-Core, a novel dual collision resolution mechanism for the four-step random access (RA) procedure of 5G communication systems. Dual-Core integrates slotted ALOHA protocol for random access preamble (RAP) transmission and a custom splitting algorithm for bandwidth request message (Msg 3) transmissions, resolving collisions in both PRACH and PUSCH. Additionally, we introduce an online controlled access class barring (ACB) to maximize the throughput of Dual-Core, with the optimal ACB rate determined by our proposed analytical model. Numerical results demonstrate that the proposed Dual-Core significantly reduces service time compared to the conventional RA procedure operating at its optimal performance, while utilizing the same amount of PUSCH resources. Furthermore, when additional PUSCH resources are available, Dual-Core efficiently utilizes them to serve more devices, a capability that conventional systems lack.

For slotted random access systems, the slotted ALOHA protocol provides the maximum throughput of 0.368 (packets/slot) while in the category of splitting (or tree) algorithms, the maximum achievable throughput can reach up to 0.487 with the first-come first-serve (FCFS) algorithm. It has been so far demonstrated that the FCFS algorithm can achieve this maximum throughput only for Poisson traffic. This may limit its application in practical systems, where packet arrivals may not be Poissonian. In this paper, we propose a novel online transmission control framework that introduces random splitting upon collisions and controls the transmission probabilities optimally at each slot by estimating the number of active users in the system. The proposed algorithm is said to be online as it estimates the number of active users slot by slot recursively, and thus can adapt to network dynamics. We first show that the splitting algorithm of our interest can achieve the throughput of 0.532 if the number of users involved in a collision could be known, which serves as a guideline for the upper limit for the random access systems with splitting. Then, when the information on the number of collided users is not available, we show that the proposed algorithm can achieve the maximum throughput of 0.487 for Poisson arrivals while achieving shorter access delay than FCFS. When more bursty traffic than Poisson process is applied, the proposed algorithm shows much better throughput and delay performance than FCFS.

For slotted random access systems with a single channel, the slotted ALOHA (S-ALOHA) protocol shows 0.368 (packets/slot) as the maximum throughput, whereas some splitting (or tree) algorithms exhibit 0.487 (packets/slot). The S-ALOHA protocol has been widely adopted even for multi-channel systems such as Long-Term Evolution (LTE), as it is more practically implementable. However, the throughput of each channel in multi-channel S-ALOHA is limited to 0.368. In order to overcome this limit and some implementational drawbacks of the existing splitting algorithms, this paper proposes a novel splitting algorithm for multi-channel slotted systems which can also adapt to dynamic system situations through estimating the number of users who have packet to transmit. We analyze the throughput of our proposed algorithm and show that the proposed algorithm is more practical than the first-come first-serve (FCFS) algorithm and shows smaller access delay than the FCFS algorithm even for a single channel system. For M-channel systems, the proposed algorithm yields the maximum throughput of 0.487M. Extensive simulations validate our analytical results.

In order to exploit the multiuser diversity in uplink non-orthogonal multiple access (NOMA) systems, we adopt the concept of the cumulative distribution function (CDF)-based scheduling for NOMA, which can fully exploit the multiuser diversity while providing fair resource sharing among the users. To support the simultaneous transmissions of multiple users with NOMA, we propose two user scheduling algorithms named CS- NOMA and CS-NOMA-Group. With CS-NOMA, in each slot the two users with the largest CDF values of their channel gains are selected. On the other hand, with CS-NOMA-Group, the users are first separated by two groups where one group consists of the users having relatively better channel gains on the average sense while the remaining users form another group. Then, in each slot, the user showing the largest CDF value of channel gain is selected in each group. For both CS-NOMA and CS- NOMA-Group, we additionally introduce a power allocation algorithm for the selected two users in each slot to balance the data rates of the two users. Through extensive simulations, we show the superior performance of CS-NOMA and CS-NOMA-Group against the conventional CS.