Professor Sarah Johnson

In this article, we consider a federated learning (FL) problem over wireless channel that takes into account the coding rate and packet transmission errors. Communication channels are modeled as packet erasure channels (PECs), where the probability of erasure is determined by block length, code rate, and signal-to-noise ratio (SNR). In spite of fluctuations in instantaneous loss of FL, we prove that the expectation of loss converges even in the presence of packet erasure. To mitigate the impact of packet erasure on FL performance, we suggest a paradigm in which the central node (CN) makes use of memory. In particular, we propose two schemes in which, in the event of packet erasure, the CN retains either the most recent local updates or the most recent global parameters. We investigate the impact of coding rate, SNR, and the CN memory on the convergence of FL. For both short- and long-packet communications, we examine a realistic scenario of a massive IoT under the assumption of error-prone transmissions. Our simulation results demonstrate that even a single memory unit has a considerable effect on the FL's efficiency in erroneous communication.

This letter studies the joint channel estimation and signal detection for the uplink power-domain non-orthogonal multiple access. The proposed technique performs both detection and estimation without the need of pilot symbols by using a clustering technique. We apply rotational-invariant coding to assist signal detection at the receiver without sending pilot symbols. We utilize Gaussian mixture model (GMM) to automatically cluster the received signals without supervision and optimize decision boundaries to improve the bit error rate (BER) performance. Simulation results show that the proposed scheme without using any pilot symbol achieves almost the same BER performance as that for the conventional maximum likelihood receiver with full channel state information.

The number of women employed in STEM in Australia is increasing, however, they continue to remain underrepresented in most industries. A significant corpus of literature on female underrepresentation has emerged in the past 20 years, however, many of those studies focus on educational access and retention and not many look at the lived experiences of women after they have left higher education. In this article, we take a different stance and explore the heterogeneous experiences of female STEM professionals in regional Australia. Through the qualitative analysis of 25 interviews, we learn what women have endured, accepted, and valued on their individual STEM journeys. While these journeys are often quite different, our interviewees independently reported having experienced similar societal prejudices and possessing similar personality traits. Our data reveals that resilience and determination proved vital for these women, as did a strong early interest in STEM. Our interviews also unearth issues in which women¿s opinions are fiercely divided, such as whether positive discrimination has been a barrier or an enabler for their careers. Based on what we have learnt from their accounts, we argue that these women have ¿survived¿ their work environments despite structural barriers, only due to their determination, resilience and fervent interest.

White matter tract (WMT) degeneration has been reported to occur following a stroke, and it is associated with post-stroke functional disturbances. White matter pathology has been suggested to be an independent predictor of post-stroke recovery. However, the factors that influence WMT remodeling are poorly understood. Cortisol is a steroid hormone released in response to prolonged stress, and elevated levels of cortisol have been reported to interfere with brain recovery. The objective of this study was to investigate the influence of corticosterone (CORT; the rodent equivalent of cortisol) on WMT structure post-stroke. Photothrombotic stroke (or sham surgery) was induced in 8-week-old male C57BL/6 mice. At 72 h, mice were exposed to standard drinking water ± CORT (100 µg/mL). After two weeks of CORT administration, mice were euthanised and brain tissue collected for histological and biochemical analysis of WMT (particularly the corpus cal-losum and corticospinal tract). CORT administration was associated with increased tissue loss within the ipsilateral hemisphere, and modest and inconsistent WMT reorganization. Further, a structural and molecular analysis of the WMT components suggested that CORT exerted effects over axons and glial cells. Our findings highlight that CORT at stress-like levels can moderately influence the reorganization and microstructure of WMT post-stroke.

Near-infrared spectroscopy (NiRS) is a relatively new technology of brain imaging with its potential in the assessment of cerebrovascular health only recently discovered. Encouraging early results suggest that NiRS can be used as an inexpensive and portable cerebrovascular health tracking device using a recently proposed pulse relaxation function (PReFx). In this paper, we propose a new NiRS timing index, TI NiRS, of cerebrovascular health. TI NiRS is a novel use of the NiRS technology. TI NiRS is motivated by the previously proved relationship of the timing of the reflected wave with vascular resistance and compliance in the context of pressure waveforms. We correlated both TI NiRS and PReFx against age, a non-exercise cardiorespiratory fitness (CRF) index, and two existing indices of cerebrovascular health, namely transcranial Doppler (TCD) augmentation index, AI TCD, and magnetic resonance imaging (MRI) blood flow pulsatility index, PI MRI. The TI NiRS correlations with Age, CRF, PI MRI and AI TCD all are significant, i.e., r= 0.53 (p= 0.002), r= - 0.44 (p= 0.011), r= 0.45 (p= 0.012) and r= 0.46 (p= 0.010), respectively. PReFx, however, did not have significant correlations with any of the vascular health factors. The proposed timing index is a reliable indicator of cerebrovascular aging factors in the NiRS waveform.

Reflected pressure waves are key to the understanding of vascular ageing, a prominent factor in major cardiovascular events. Several different metrics have been proposed to index the effect of wave reflection on the pressure waveform and thereby serve as an indicator of vascular ageing. The extent to which these indices are influenced by factors other than vascular health remains a matter of concern. In this paper, we use transmission-line theory to derive a mathematical model for the reflection time (Trefl), and the augmentation index (AI), assuming a general extended model of the arterial system. Then, we test the proposed model against values reported in the literature. Finally, we discuss insights from the model to common observations in the literature such as age-related ¿shift¿ in the reflection site, the variation of AI with heart rate, and the flattening of Trefl in older participants.

Uplink power domain non-orthogonal multiple access (PD-NOMA) allows multiple users to simultaneously transmit their data over the same resource block by using different power levels, which are exploited by the base station (BS) for multi-user detection (MUD). Conventional PD-NOMA receivers can perform MUD when the data streams from all the multiplexed users are continuous, and the received signal is always a superposition of symbols from each user. However, the data received from each user may not always be continuous due to sporadic activity, and may contain missed packets or moments of silence. In such scenarios, the BS needs to precisely identify the users whose data packets are received at a particular time to successfully recover their messages. In this context, this letter proposes multiple detectors to efficiently detect user activity and perform MUD in such situations. It is shown through simulations that the proposed detectors significantly outperform conventional PD-NOMA receivers in terms of the bit error rate.

Transcranial Doppler (TCD) blood flow velocity has been extensively used in biomedical research as it provides a cost-effective and relatively simple approach to assess changes in cerebral blood flow dynamics and track cerebrovascular health status. In this article we introduce a new TCD-based timing index, TITCD, as an indicator of vascular stiffening and vascular health. We investigate the correlations of the new index and the existing indices, namely the pulsatility index and the augmentation index, with age, cardiorespiratory fitness (CRF) and magnetic resonance imaging (MRI) blood flow pulsatility index (PIMRI). Notably, the new index showed stronger correlations with CRF (r = -0.79) and PIMRI (r = 0.53) compared with the augmentation index (r = -0.65 with CRF and no significant correlation with PIMRI) and the pulsatility index (no significant correlations with CRF or PIMRI), and a similar correlation with age as the augmentation index. The clearer relationship of the proposed timing index with vascular aging factors underlines its utility as an early indicator of vascular stiffening.

Secondary neurodegeneration (SND) is an insidious and progressive condition involving the death of neurons in regions of the brain that were connected to but undamaged by the initial stroke. Our group have published compelling evidence that exposure to psychological stress can significantly exacerbate the severity SND, a finding that has considerable clinical implications given that stroke-survivors often report experiencing high and unremitting levels of psychological stress. It may be possible to use one or more targeted pharmacological approaches to limit the negative effects of stress on the recovery process but in order to move forward with this approach the most critical stress signals have to be identified. Accordingly, in the current study we have directed our attention to examining the potential effects of corticosterone, delivered orally at stress-like levels. Our interest is to determine how similar the effects of corticosterone are to stress on repair and remodelling that is known to occur after stroke. The study involved 4 groups, sham and stroke, either administered corticosterone or normal drinking water. The functional impact was assessed using the cylinder task for paw asymmetry, grid walk for sensorimotor function, inverted grid for muscle strength and coordination and open field for anxiety-like behaviour. Biochemically and histologically, we considered disturbances in main cellular elements of the neurovascular unit, including microglia, astrocytes, neurons and blood vessels using both immunohistochemistry and western blotting. In short, we identified that corticosterone delivery after stroke results in significant suppression of key microglial and astroglial markers. No changes were observed on the vasculature and in neuronal specific markers. No changes were identified for sensorimotor function or anxiety-like behaviour. We did, however, observe a significant change in motor function as assessed using the inverted grid walk test. Collectively, these results suggest that pharmacologically targeting corticosterone levels in the future may be warranted but that such an approach is unlikely to limit all the negative effects associated with exposure to chronic stress.

Exposure to psychological stress is known to seriously disrupt the operation of the substantia nigra (SN) and may in fact initiate the loss of dopaminergic neurons within the SN. In this study, we aimed to investigate how chronic stress modified the SN in adult male mice. Using a paradigm of repeated restraint stress (an average of 20¿h per week for 6¿weeks), we examined changes within the SN using western blotting and immunohistochemistry. We demonstrated that chronic stress was associated with a clear loss of dopaminergic neurons within the SN. The loss of dopaminergic neurons was accompanied by higher levels of oxidative stress damage, indexed by levels of protein carbonylation and strong suppression of both microglial and astrocytic responses. In addition, we demonstrated for the first time, that chronic stress alone enhanced the aggregation of a-synuclein into the insoluble protein fraction. These results indicate that chronic stress triggered loss of dopaminergic neurons by increasing oxidative stress, suppressing glial neuroprotective functions and enhancing the aggregation of the neurotoxic protein, a-synuclein. Collectively, these results reinforce the negative effects of chronic stress on the viability of dopaminergic cells within the SN.

Exposure to chronic stress following stroke has been shown, both clinically and pre-clinically, to impact negatively on the recovery process. While this phenomenon is well established, the specific mechanisms involved have remained largely unexplored. One obvious signaling pathway through which chronic stress may impact on the recovery process is via corticosterone, and its effects on microglial activity and vascular remodeling. In the current study, we were interested in examining how orally delivered corticosterone at a stress-like concentration impacted on microglial activity and vascular remodeling after stroke. We identified that corticosterone administration for two weeks following stroke significantly increased tissue loss and decreased the weight of the spleen and thymus. We also identified that corticosterone administration significantly altered the expression of the key microglial complement receptor, CD11b after stroke. Corticosterone administration did not alter the expression of the vessel basement membrane protein, Collagen IV after stroke. Together, these results suggest that corticosterone is likely to represent only one of the major stress signals responsible for driving the negative impacts of chronic stress on recovery.

Stroke induces tissue death both at the site of infarction and at secondary sites connected to the primary infarction. This latter process has been referred to as secondary neurodegeneration (SND). Using predominantly fixed tissue analyses, microglia have been implicated in regulating the initial response at both damage sites post-stroke. In this study, we used acute slice based multiphoton imaging, to investigate microglia dynamic process movement in mice 14 days after a photothrombotic stroke. We evaluated the baseline motility and process responses to locally induced laser damage in both the peri-infarct (PI) territory and the ipsilateral thalamus, a major site of post-stroke SND. Our findings show that microglia process extension toward laser damage within the thalamus is lost, yet remains robustly intact within the PI territory. However, microglia at both sites displayed an activated morphology and elevated levels of commonly used activation markers (CD68, CD11b), indicating that the standardly used fixed tissue metrics of microglial ¿activity¿ are not necessarily predictive of microglia function. Analysis of the purinergic P2Y12 receptor, a key regulator of microglia process extension, revealed an increased somal localization on nonresponsive microglia in the thalamus. To our knowledge, this is the first study to identify a non-responsive microglia phenotype specific to areas of SND post-stroke, which cannot be identified by the classical assessment of microglia activation but rather the localization of P2Y12 to the soma.

Sickness behaviors have become the focus of great interest in recent years as they represent a clear case of how peripheral disturbances in immune signaling can disrupt quite complex behaviors. In the current study, we were interested in examining whether we could identify any significant morphological disturbances in microglia associated with these sickness-like behaviors in adult male Sprague-Dawley rats. We chose lipopolysaccharide (LPS 100 µg/kg/i.p.), to induce sickness-like behaviors as it is the most well-validated approach to do so in rodents and humans. We were particularly interested in examining changes in microglia within the prefrontal cortex (PFC) as several recent neuroimaging studies have highlighted significant functional changes in this region following peripheral LPS administration. Paraformaldehyde-fixed tissue was collected from animals 24 h post LPS administration and labeled immunohistochemically with an antibody directed to bind to Iba-1, a protein known to be involved in the structural remodeling of microglia. To analyze changes, we have made use of two recently described image analysis procedures. The first is known as cumulative threshold spectra (CTS) analysis. The second involves the unsupervised digital reconstruction of microglia. We undertook these complementary analysis of microglial cells in the both the pre- and infralimbic divisions of the PFC. Our results indicated that microglial soma size was significantly enlarged, while cell processes had contracted slightly following LPS administration. To our knowledge this study is to first to definitely demonstrate substantial microglial disturbances within the PFC following LPS delivered at a dose that was sufficient to induce significant sickness-like behavior.

This paper investigates the design of low-complexity error correction codes for the verification step in continuous variable quantum key distribution (CVQKD) systems. We design new coding schemes based on quasi-cyclic repeat-accumulate codes which demonstrate good performances for CVQKD reconciliation. Given quasi-cyclic repeat-accumulate codes' commercial maturity, low implementation complexity and existing high-speed ASIC implementations, this makes these codes viable candidates for commercial CVQKD systems.

A source coding problem over a noiseless broadcast channel where the source is preinformed about the contents of the cache of all receivers, is an index coding problem. Furthermore, if each message is requested by one receiver, then we call this an index coding problem with a unicast message setting. This problem can be represented by a directed graph. In this paper, we first define a structure (we call generalized interlinked cycle (GIC)) in directed graphs. A GIC consists of cycles which are interlinked in some manner (i.e., not disjoint), and it turns out that the GIC is a generalization of cliques and cycles. We then propose a simple scalar linear encoding scheme with linear time encoding complexity. This scheme exploits GICs in the digraph. We prove that our scheme is optimal for a class of digraphs with message packets of any length. Moreover, we show that our scheme can outperform existing techniques, e.g., partial clique cover, local chromatic number, composite-coding, and interlinked cycle cover.

Intrinsic Optical Signal imaging is a technique which allows the visualisation and mapping of activity related changes within the brain with excellent spatial and temporal resolution. We analysed a variety of signal and image processing techniques applied to real mouse imaging data. The results were compared in an attempt to overcome the unique issues faced when performing the technique on mice and improve the understanding of post processing options available.

We study the index coding problem in the unicast message setting, i.e., where each message is requested by one unique receiver. This problem can be modeled by a directed graph. We propose a new scheme called interlinked cycle cover, which exploits interlinked cycles in the directed graph, for designing index codes. This new scheme generalizes the existing clique cover and cycle cover schemes. We prove that for a class of infinitely many digraphs with messages of any length, interlinked cycle cover provides an optimal index code. Furthermore, the index code is linear with linear time encoding complexity.

This paper investigates the capacity regions of two-receiver broadcast channels where each receiver (i) has both common and private-message requests, and (ii) knows part of the private message requested by the other receiver as side information. We first propose a transmission scheme and derive an inner bound for the two-receiver memoryless broadcast channel. We next prove that this inner bound is tight for the deterministic channel and the more capable channel, thereby establishing their capacity regions.We show that this inner bound is also tight for all classes of two-receiver broadcast channels whose capacity regions were known prior to this work. Our proposed scheme is consequently a unified capacity-achieving scheme for these classes of broadcast channels.

This paper considers the finite-field multi-way relay channel with pairwise message sharing, where multiple users exchange messages through a single relay and where the users may share parts of their source messages (meaning that some message parts are known/common to more than one user). In this paper, we design an optimal functional-decode-forward coding scheme that takes the shared messages into account. More specifically, we design an optimal function for the relay to decode (from the users on the uplink) and forward (back to the users on the downlink). We then show that this proposed function-decode-forward coding scheme can achieve the capacity region of the finite-field multi-way relay channel with pairwise message sharing. This paper generalizes our previous result for the case of three users to any number of users. © 2014 IEEE.

This paper investigates the capacity region of three-receiver AWGN broadcast channels where the receivers (i) have private-message requests and (ii) know the messages requested by some other receivers as side information. We classify these channels based on their side information into eight groups, and construct different transmission schemes for the groups. For six groups, we characterize the capacity region, and show that it improves both the best known inner and outer bounds. For the remaining two groups, we improve the best known inner bound by using side information during channel decoding at the receivers. © 2014 IEEE.

In this paper low-density parity-check (LDPC) codes are considered for burst erasure channels. Multiple burst erasure correcting LDPC codes are constructed using superposition and the burst erasure correcting performances of the resulting codes are derived as a property of the stopping set size of the base matrices and the choice of permutation matrices for the superposition.

We investigate the effect of channel gain quantization on the information capacity of unknown time varying flat fading channels. The phase and/or amplitude of the flat fading channel gain is modelled as a finite state Markov (FSM) process and the information capacity of the FSM channel is calculated numerically as a measure for choosing the number of channel quantization levels, as well as quantization thresholds. The results indicate that for binary signalling, the capacity is saturated beyond 8 to 16 levels of phase and 8 to 16 levels of amplitude quantization.